1 00:00:01,120 --> 00:00:03,490 The following content is provided under a Creative 2 00:00:03,490 --> 00:00:04,880 Commons license. 3 00:00:04,880 --> 00:00:07,090 Your support will help MIT OpenCourseWare 4 00:00:07,090 --> 00:00:11,180 continue to offer high quality educational resources for free. 5 00:00:11,180 --> 00:00:13,750 To make a donation or to view additional materials 6 00:00:13,750 --> 00:00:17,680 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:17,680 --> 00:00:18,586 at ocw.mit.edu. 8 00:00:23,070 --> 00:00:25,500 GARY GENSLER: All right, so we're going to try to-- 9 00:00:25,500 --> 00:00:28,450 I was giving a little bit more time for more people 10 00:00:28,450 --> 00:00:29,910 to show up-- 11 00:00:29,910 --> 00:00:32,070 chat about what did we cover this semester. 12 00:00:32,070 --> 00:00:35,010 So I'm going to try to talk about-- 13 00:00:35,010 --> 00:00:36,920 we started out-- money and ledgers, 14 00:00:36,920 --> 00:00:41,520 and why does that matter, and if you go away from all this. 15 00:00:41,520 --> 00:00:47,820 Satoshi Nakamoto's innovation-- what is that again? 16 00:00:47,820 --> 00:00:50,250 The economics of blockchain technology, 17 00:00:50,250 --> 00:00:54,400 why the financial sector is so entwined in all of this, 18 00:00:54,400 --> 00:00:56,780 though it's not the only place we're going. 19 00:00:56,780 --> 00:00:59,790 A little bit about crypto finance and public policy 20 00:00:59,790 --> 00:01:04,680 frameworks, and then wrap it up with a quote from Ben 21 00:01:04,680 --> 00:01:07,620 Franklin-- will be the last thing we talk about at the end, 22 00:01:07,620 --> 00:01:08,790 about paying it forward. 23 00:01:11,430 --> 00:01:13,020 I don't know how many people read 24 00:01:13,020 --> 00:01:16,690 the article I wrote for Coin Desk, which 25 00:01:16,690 --> 00:01:18,520 will run in a day or two or something, 26 00:01:18,520 --> 00:01:27,430 but that, too, just like helped me wrap up the whole semester. 27 00:01:27,430 --> 00:01:30,670 As I told you, I'm neither a maximalist or a minimalist, 28 00:01:30,670 --> 00:01:32,260 but in a few minutes, I'm probably 29 00:01:32,260 --> 00:01:37,570 going to ask you all to kind of let me know where you ended up. 30 00:01:37,570 --> 00:01:38,860 So the role of money. 31 00:01:38,860 --> 00:01:41,800 Anybody remember the role of money 32 00:01:41,800 --> 00:01:43,570 in an economy, the three things? 33 00:01:46,432 --> 00:01:48,730 AUDIENCE: Medium of exchange, unit of account, 34 00:01:48,730 --> 00:01:49,990 store of value. 35 00:01:49,990 --> 00:01:52,360 GARY GENSLER: Medium of exchange, unit of account, 36 00:01:52,360 --> 00:01:53,350 store of value. 37 00:01:55,900 --> 00:01:57,910 I don't think we really know, when 38 00:01:57,910 --> 00:02:01,360 we go through all the history and the archeology of it, 39 00:02:01,360 --> 00:02:05,660 which came first, but all three really matter. 40 00:02:05,660 --> 00:02:09,160 So how many people think that Bitcoin-- 41 00:02:09,160 --> 00:02:10,060 just Bitcoin. 42 00:02:10,060 --> 00:02:12,940 I'm not talking about the other 1,600 tokens-- 43 00:02:12,940 --> 00:02:20,290 Bitcoin fulfills these three roles of money? 44 00:02:20,290 --> 00:02:21,635 James, was that a hand up? 45 00:02:21,635 --> 00:02:22,260 AUDIENCE: Yeah. 46 00:02:22,260 --> 00:02:23,077 Yep. 47 00:02:23,077 --> 00:02:24,410 GARY GENSLER: You're saying yes. 48 00:02:24,410 --> 00:02:25,815 AUDIENCE: Yes. 49 00:02:25,815 --> 00:02:26,690 GARY GENSLER: Alexis? 50 00:02:26,690 --> 00:02:27,420 Oh, we got two. 51 00:02:27,420 --> 00:02:27,920 Hugo? 52 00:02:30,960 --> 00:02:33,090 Somebody take the other side. 53 00:02:33,090 --> 00:02:33,780 Tom? 54 00:02:33,780 --> 00:02:35,060 AUDIENCE: Nope. 55 00:02:35,060 --> 00:02:36,060 GARY GENSLER: That's it? 56 00:02:36,060 --> 00:02:36,560 That's it? 57 00:02:36,560 --> 00:02:38,765 A whole semester, and all you can say is nope? 58 00:02:38,765 --> 00:02:39,390 AUDIENCE: Yeah. 59 00:02:39,390 --> 00:02:39,750 GARY GENSLER: Nope? 60 00:02:39,750 --> 00:02:40,610 All right. 61 00:02:40,610 --> 00:02:42,990 Anybody want to give more? 62 00:02:42,990 --> 00:02:44,446 Brotish? 63 00:02:44,446 --> 00:02:47,740 AUDIENCE: So not a store of value because of [INAUDIBLE].. 64 00:02:47,740 --> 00:02:50,280 GARY GENSLER: So Brotish says not a store of value, 65 00:02:50,280 --> 00:02:53,310 but isn't it $60 billion of value right now? 66 00:02:53,310 --> 00:02:54,990 AUDIENCE: [INAUDIBLE]. 67 00:02:54,990 --> 00:02:58,610 So there is a value, but it's too volatile. 68 00:02:58,610 --> 00:02:59,775 GARY GENSLER: James? 69 00:02:59,775 --> 00:03:03,210 AUDIENCE: I respectfully disagree. 70 00:03:03,210 --> 00:03:06,360 It's volatile when you think of it in exchange of a dollar, 71 00:03:06,360 --> 00:03:08,940 but if you move your mindset, a bitcoin 72 00:03:08,940 --> 00:03:10,290 is a bitcoin is a bitcoin. 73 00:03:10,290 --> 00:03:10,930 Right? 74 00:03:10,930 --> 00:03:15,818 It's only volatile if you think [INAUDIBLE] a dollar. 75 00:03:15,818 --> 00:03:16,860 GARY GENSLER: Oh, my god. 76 00:03:16,860 --> 00:03:19,050 Here we go. 77 00:03:19,050 --> 00:03:20,480 Ross? 78 00:03:20,480 --> 00:03:23,005 AUDIENCE: But that same thing is true of corn, right? 79 00:03:23,005 --> 00:03:24,680 A bushel of corn is-- 80 00:03:24,680 --> 00:03:25,722 I'm with Brotish on this. 81 00:03:25,722 --> 00:03:27,097 GARY GENSLER: I'm sorry, were you 82 00:03:27,097 --> 00:03:28,920 in the no camp or the yes camp? 83 00:03:28,920 --> 00:03:30,150 That it's a store of value. 84 00:03:30,150 --> 00:03:31,230 AUDIENCE: I'm in the no camp. 85 00:03:31,230 --> 00:03:31,938 GARY GENSLER: No? 86 00:03:31,938 --> 00:03:33,240 Is it a medium of exchange? 87 00:03:33,240 --> 00:03:34,050 AUDIENCE: Yes. 88 00:03:34,050 --> 00:03:35,290 GARY GENSLER: Oh, medium of exchange. 89 00:03:35,290 --> 00:03:35,800 Sean? 90 00:03:35,800 --> 00:03:39,895 AUDIENCE: In response to [INAUDIBLE] none 91 00:03:39,895 --> 00:03:41,520 of the currency can actually be treated 92 00:03:41,520 --> 00:03:42,670 as a stand alone currency. 93 00:03:42,670 --> 00:03:45,780 Everything has to be treated as a pair. 94 00:03:45,780 --> 00:03:48,460 So everything's a relative comparison. 95 00:03:48,460 --> 00:03:51,720 So you can't really treat Bitcoin on a standalone basis. 96 00:03:55,457 --> 00:03:57,290 AUDIENCE: I don't know if I agree with that. 97 00:03:57,290 --> 00:04:00,180 But I would say that pretty much anything can fit into all three 98 00:04:00,180 --> 00:04:01,000 of these buckets. 99 00:04:01,000 --> 00:04:03,430 It just depends on how dependable it is. 100 00:04:03,430 --> 00:04:05,700 In those three buckets, with US dollar's 101 00:04:05,700 --> 00:04:07,710 very good at being medium of exchange, 102 00:04:07,710 --> 00:04:09,690 very good at being a dependable store of value, 103 00:04:09,690 --> 00:04:11,490 and a steady unit of account. 104 00:04:11,490 --> 00:04:16,867 But you could also have a bushel of wheat or something. 105 00:04:16,867 --> 00:04:18,450 It would be all three of these things, 106 00:04:18,450 --> 00:04:19,320 but it wouldn't be equality. 107 00:04:19,320 --> 00:04:20,550 GARY GENSLER: Right, but-- 108 00:04:20,550 --> 00:04:21,519 please, Jihee. 109 00:04:21,519 --> 00:04:25,860 AUDIENCE: So the fiat currency-- the reason why 110 00:04:25,860 --> 00:04:27,910 I think US dollar is stable is because there 111 00:04:27,910 --> 00:04:31,650 is a US central bank that actually backs up that. 112 00:04:31,650 --> 00:04:33,450 Because it's just that fiat currency. 113 00:04:33,450 --> 00:04:36,900 It is a liability of the central bank, 114 00:04:36,900 --> 00:04:40,998 whereas Bitcoin doesn't have a center 115 00:04:40,998 --> 00:04:42,040 to worry that banks have. 116 00:04:42,040 --> 00:04:43,774 GARY GENSLER: Nice music that we're getting. 117 00:04:43,774 --> 00:04:44,710 [INAUDIBLE] a serenade. 118 00:04:44,710 --> 00:04:46,252 AUDIENCE: So therefore, I don't think 119 00:04:46,252 --> 00:04:48,363 the store of value argument will work. 120 00:04:48,363 --> 00:04:49,530 GARY GENSLER: So you think-- 121 00:04:49,530 --> 00:04:50,905 [INAUDIBLE] saying, well, there's 122 00:04:50,905 --> 00:04:54,310 no central bank behind it, so maybe it doesn't hold up. 123 00:04:54,310 --> 00:04:57,120 But if you take anything away from the class-- 124 00:04:57,120 --> 00:05:00,630 and I'm going to have some other takeaways, too. 125 00:05:00,630 --> 00:05:04,470 Remember, the three things that about the role of money-- 126 00:05:04,470 --> 00:05:06,870 but it's a social construct. 127 00:05:06,870 --> 00:05:08,430 Even with a central bank-- 128 00:05:08,430 --> 00:05:11,970 a central bank is, ultimately-- we enshrined in law. 129 00:05:11,970 --> 00:05:15,420 We enshrine it in a big institution, and bricks, 130 00:05:15,420 --> 00:05:18,770 and mortar, and columns usually. 131 00:05:18,770 --> 00:05:21,520 You always have to have those columns. 132 00:05:21,520 --> 00:05:24,750 But it's still a social construct. 133 00:05:24,750 --> 00:05:32,160 And Bitcoin does, in some ways, have all three of these. 134 00:05:32,160 --> 00:05:34,260 It's just not broadly acceptable. 135 00:05:34,260 --> 00:05:37,030 It's rarely used as a unit of account, 136 00:05:37,030 --> 00:05:39,390 but it is sometimes used as a unit account 137 00:05:39,390 --> 00:05:42,100 in some initial coin offerings. 138 00:05:42,100 --> 00:05:45,180 It's not used, generally, as a medium of exchange. 139 00:05:45,180 --> 00:05:47,460 But some people are paid in Bitcoin some. 140 00:05:47,460 --> 00:05:53,010 Software developers are actually paid in Bitcoin. 141 00:05:53,010 --> 00:05:54,030 And it's $60 billion. 142 00:05:54,030 --> 00:05:55,440 It might be volatile. 143 00:05:55,440 --> 00:05:57,460 It has a store value. 144 00:05:57,460 --> 00:06:00,900 So I'm not trying to take maximalist or minimalist, 145 00:06:00,900 --> 00:06:03,660 but I'm probably between Tom and James, 146 00:06:03,660 --> 00:06:07,290 because it does have all of those qualities 147 00:06:07,290 --> 00:06:12,390 in to some extent in there. 148 00:06:12,390 --> 00:06:16,440 Early money-- remember some of our walk down memory lane. 149 00:06:16,440 --> 00:06:22,620 Some of these early monies fell apart, became extinct, 150 00:06:22,620 --> 00:06:24,428 when they got debased. 151 00:06:24,428 --> 00:06:26,220 And we talked about, early in the semester, 152 00:06:26,220 --> 00:06:27,900 even the story about it. 153 00:06:27,900 --> 00:06:30,990 It was, I think, a British-- 154 00:06:30,990 --> 00:06:33,960 when the British got to the island of Yap, 155 00:06:33,960 --> 00:06:36,900 they went off a couple miles and quarried the stones 156 00:06:36,900 --> 00:06:39,450 and kept bringing more yap stones there. 157 00:06:39,450 --> 00:06:42,240 And all of a sudden, it got debased. 158 00:06:42,240 --> 00:06:45,710 And then, of course, money turned to paper money. 159 00:06:45,710 --> 00:06:47,950 And in the bottom left-hand corner, 160 00:06:47,950 --> 00:06:50,730 it started as warehouse receipts. 161 00:06:50,730 --> 00:06:53,040 So the paper was just a representation 162 00:06:53,040 --> 00:06:54,630 of the store of value-- 163 00:06:54,630 --> 00:06:59,310 the metal, or the corn, or the wheat that we had earlier. 164 00:06:59,310 --> 00:07:02,460 Another important thing was ledgers. 165 00:07:02,460 --> 00:07:04,930 Who wants to remind the class what a ledger is? 166 00:07:04,930 --> 00:07:08,510 Why do ledgers matter? 167 00:07:08,510 --> 00:07:10,310 Anybody want to-- ledgers? 168 00:07:10,310 --> 00:07:13,865 All right, clearly I didn't make it on ledgers. 169 00:07:16,790 --> 00:07:18,605 Hugo? 170 00:07:18,605 --> 00:07:21,780 AUDIENCE: It's transaction history, basically. 171 00:07:21,780 --> 00:07:23,630 GARY GENSLER: A store transaction history. 172 00:07:23,630 --> 00:07:26,620 What else is there? 173 00:07:26,620 --> 00:07:27,460 AUDIENCE: Balances. 174 00:07:27,460 --> 00:07:29,312 GARY GENSLER: Balances of? 175 00:07:29,312 --> 00:07:30,640 AUDIENCE: Accounts. 176 00:07:30,640 --> 00:07:31,810 GARY GENSLER: Accounts. 177 00:07:31,810 --> 00:07:34,420 So it stores a balance and a transaction. 178 00:07:34,420 --> 00:07:37,980 So you can think of it as a flow and a balance. 179 00:07:37,980 --> 00:07:40,510 The flow is the transaction-- 180 00:07:40,510 --> 00:07:43,380 I give Hugo money. 181 00:07:43,380 --> 00:07:46,470 Regardless of who gives Hugo money, here's his balance. 182 00:07:46,470 --> 00:07:49,410 So an income statement and a balance sheet 183 00:07:49,410 --> 00:07:51,210 is a flow and a balance. 184 00:07:51,210 --> 00:07:57,440 But ledgers and keeping ledgers go back thousands of years. 185 00:07:57,440 --> 00:08:00,550 So blockchain technology is really about money, 186 00:08:00,550 --> 00:08:02,860 but it's also about a database that has ledgers. 187 00:08:02,860 --> 00:08:04,750 And ledgers usually store things of value. 188 00:08:07,270 --> 00:08:09,130 So then we got to fiat currency. 189 00:08:09,130 --> 00:08:11,460 And fiat currency, as Jihee said, 190 00:08:11,460 --> 00:08:14,440 they were represented by central bank notes, central bank 191 00:08:14,440 --> 00:08:16,285 reserves, and bank deposits. 192 00:08:16,285 --> 00:08:17,800 So there's three things. 193 00:08:17,800 --> 00:08:20,950 Does anybody want to see if anybody 194 00:08:20,950 --> 00:08:22,480 listened in those classes? 195 00:08:22,480 --> 00:08:24,130 [LAUGHS] I'm kidding around. 196 00:08:24,130 --> 00:08:26,680 Elan, you're going to say, why is it those three things? 197 00:08:26,680 --> 00:08:29,872 They're all forms of fiat money. 198 00:08:29,872 --> 00:08:31,330 AUDIENCE: I thought you were asking 199 00:08:31,330 --> 00:08:34,950 about the illicit activity in that list. 200 00:08:34,950 --> 00:08:35,980 Is that the question? 201 00:08:35,980 --> 00:08:37,480 GARY GENSLER: All right, answer the question 202 00:08:37,480 --> 00:08:38,340 you want to answer. 203 00:08:38,340 --> 00:08:38,965 AUDIENCE: Yeah. 204 00:08:38,965 --> 00:08:39,960 [LAUGHTER] 205 00:08:39,960 --> 00:08:41,335 [INAUDIBLE] there are things that 206 00:08:41,335 --> 00:08:45,466 are important in fiat currency, or is helpful. 207 00:08:45,466 --> 00:08:50,440 Fiat currency is important for paying taxes. 208 00:08:50,440 --> 00:08:56,230 And it is allegedly backed by the social understanding 209 00:08:56,230 --> 00:09:00,707 that the central bank will respect those notes at the end. 210 00:09:00,707 --> 00:09:01,540 GARY GENSLER: Right. 211 00:09:01,540 --> 00:09:03,850 So it's actually-- you're on to this. 212 00:09:03,850 --> 00:09:06,130 It's accepted for taxes. 213 00:09:06,130 --> 00:09:07,330 It's legal tender. 214 00:09:07,330 --> 00:09:09,283 So actually, society gets together, 215 00:09:09,283 --> 00:09:11,200 through its parliament, through its executive, 216 00:09:11,200 --> 00:09:13,740 and passes a law that says it's legal tender. 217 00:09:13,740 --> 00:09:15,280 That gives it a huge leg up. 218 00:09:15,280 --> 00:09:17,440 But this is what it's really about. 219 00:09:17,440 --> 00:09:19,900 It has huge network effects. 220 00:09:19,900 --> 00:09:21,890 And I wrote this in the CoinDesk article. 221 00:09:21,890 --> 00:09:26,440 The question I have for any new currency 222 00:09:26,440 --> 00:09:28,780 is, how does it compete with those incredible network 223 00:09:28,780 --> 00:09:30,250 effects? 224 00:09:30,250 --> 00:09:33,400 And we talk about Facebook has incredible network effects 225 00:09:33,400 --> 00:09:34,510 because-- 226 00:09:34,510 --> 00:09:36,400 two billion members. 227 00:09:36,400 --> 00:09:39,880 But even money, as a technology, has a network effect 228 00:09:39,880 --> 00:09:43,090 because people readily accept it as a unit of account, 229 00:09:43,090 --> 00:09:47,480 a medium of exchange, a store of value. 230 00:09:47,480 --> 00:09:51,200 And it only loses that network effect, usually, 231 00:09:51,200 --> 00:09:54,050 when it gets debased, when people 232 00:09:54,050 --> 00:09:56,840 can't be confident that it's a good store of value 233 00:09:56,840 --> 00:09:59,240 because somebody else can mint a lot of it, 234 00:09:59,240 --> 00:10:06,170 or print a lot of it, or a king starts to print too much of it. 235 00:10:06,170 --> 00:10:09,380 But it has an extraordinary network effect. 236 00:10:09,380 --> 00:10:11,030 On the other hand, sometimes currencies 237 00:10:11,030 --> 00:10:13,850 are a little bit challenged, like the euro right now. 238 00:10:13,850 --> 00:10:15,930 Will the euro ultimately work? 239 00:10:15,930 --> 00:10:18,230 That's for a different class and a different lecture, 240 00:10:18,230 --> 00:10:21,650 but will the euro ultimately work long term? 241 00:10:21,650 --> 00:10:26,300 I have my doubts whether it will work for decades to come, 242 00:10:26,300 --> 00:10:29,620 because there's so many jurisdictions. 243 00:10:29,620 --> 00:10:32,230 And there's not a unified fiscal account, 244 00:10:32,230 --> 00:10:40,790 not a unified economy, and not truly free movement of labor 245 00:10:40,790 --> 00:10:44,850 across that continent. 246 00:10:44,850 --> 00:10:46,590 So that's fiat currency. 247 00:10:46,590 --> 00:10:50,970 It's sort of the game in town that then Satoshi Nakamoto 248 00:10:50,970 --> 00:10:53,070 is was addressing. 249 00:10:53,070 --> 00:10:55,900 And you all read the eight-page paper, 250 00:10:55,900 --> 00:10:57,540 but that was the first line. 251 00:10:57,540 --> 00:10:59,218 "Been working on a new electronic cash 252 00:10:59,218 --> 00:11:01,260 system that's fully peer-to-peer, with no trusted 253 00:11:01,260 --> 00:11:02,280 third party." 254 00:11:02,280 --> 00:11:04,470 Basically, as Jihee said, no central bank. 255 00:11:04,470 --> 00:11:06,450 Forget about that central bank. 256 00:11:06,450 --> 00:11:12,000 That was the core inspiration that was there. 257 00:11:12,000 --> 00:11:15,180 But interestingly, it wasn't the first try. 258 00:11:15,180 --> 00:11:17,160 And we talked early in the semester 259 00:11:17,160 --> 00:11:20,730 about at least a handful of early attempts 260 00:11:20,730 --> 00:11:24,400 that all failed because they were still 261 00:11:24,400 --> 00:11:29,320 centralized-- like DigiCash was still very centralized. 262 00:11:29,320 --> 00:11:31,870 Or they couldn't solve the double spending. 263 00:11:31,870 --> 00:11:34,510 On the internet, you can send an email twice. 264 00:11:34,510 --> 00:11:37,090 It doesn't need a central authority. 265 00:11:37,090 --> 00:11:38,620 So what if it gets copied? 266 00:11:38,620 --> 00:11:42,140 But the double spending is the big issue. 267 00:11:42,140 --> 00:11:47,940 So how did he solve it, or she solve it? 268 00:11:47,940 --> 00:11:52,060 Does anybody still not know who Satoshi Nakamoto is? 269 00:11:52,060 --> 00:11:52,560 No? 270 00:11:52,560 --> 00:11:53,820 All right. 271 00:11:53,820 --> 00:11:55,290 I was hoping, I was hoping. 272 00:11:57,980 --> 00:12:04,160 So we talked about blocks of data 273 00:12:04,160 --> 00:12:08,655 using cryptography and consensus. 274 00:12:08,655 --> 00:12:10,280 So if you're at a dinner party, and you 275 00:12:10,280 --> 00:12:13,767 can't remember anything else from this class, [LAUGHS] 276 00:12:13,767 --> 00:12:15,350 maybe just remember, oh, yeah, there's 277 00:12:15,350 --> 00:12:19,570 something about blocks of data using a bunch of cryptography 278 00:12:19,570 --> 00:12:21,560 and consensus. 279 00:12:21,560 --> 00:12:22,910 These slides are on Canvas. 280 00:12:22,910 --> 00:12:26,350 Download this one slide, you're the expert. 281 00:12:26,350 --> 00:12:28,745 Neha Nerula actually created this slide, 282 00:12:28,745 --> 00:12:32,060 and I've been using it a lot. 283 00:12:32,060 --> 00:12:35,420 But anybody remember who invented this whole concept 284 00:12:35,420 --> 00:12:36,550 of blocks of data? 285 00:12:39,970 --> 00:12:40,470 James? 286 00:12:40,470 --> 00:12:41,640 AUDIENCE: Adam someone? 287 00:12:41,640 --> 00:12:42,140 Back? 288 00:12:42,140 --> 00:12:43,765 GARY GENSLER: Adam Back is a good name, 289 00:12:43,765 --> 00:12:46,140 but he's not the guy that invented this. 290 00:12:46,140 --> 00:12:49,972 He invented something else, which was proof-of-work. 291 00:12:52,690 --> 00:12:54,600 Timestamp append-only logs-- what's 292 00:12:54,600 --> 00:12:57,840 the longest running blockchain in the world? 293 00:12:57,840 --> 00:12:59,100 AUDIENCE: New York Times? 294 00:12:59,100 --> 00:12:59,655 GARY GENSLER: New York Times. 295 00:12:59,655 --> 00:13:00,600 [INTERPOSING VOICES] 296 00:13:00,600 --> 00:13:01,433 AUDIENCE: Crossword. 297 00:13:01,433 --> 00:13:05,700 GARY GENSLER: All right, here he is. 298 00:13:05,700 --> 00:13:10,480 So two Bell Labs scientists-- this is one of them. 299 00:13:10,480 --> 00:13:14,560 Two Bell Labs scientists came up with this concept 300 00:13:14,560 --> 00:13:19,000 that you could take a bunch of data, put it through what's 301 00:13:19,000 --> 00:13:23,320 called a cryptographic hash function, 302 00:13:23,320 --> 00:13:24,637 and that's a commitment scheme. 303 00:13:24,637 --> 00:13:26,970 You couldn't change it, because if you changed anything, 304 00:13:26,970 --> 00:13:30,360 the hash function would change. 305 00:13:30,360 --> 00:13:32,430 So what's a cryptographic hash function? 306 00:13:32,430 --> 00:13:34,350 That's when I use this idea of you 307 00:13:34,350 --> 00:13:37,080 can take the entire library of Congress, 308 00:13:37,080 --> 00:13:39,780 and put it into a little thing called a hash function, 309 00:13:39,780 --> 00:13:45,150 and end up with a string of characters, 310 00:13:45,150 --> 00:13:52,280 a hexadecimal output. 311 00:13:52,280 --> 00:13:59,790 But a hash function is the key to what this guy figured out. 312 00:13:59,790 --> 00:14:02,660 And, of course, this. 313 00:14:02,660 --> 00:14:04,370 That's the key. 314 00:14:04,370 --> 00:14:06,890 There was only two bits of cryptography 315 00:14:06,890 --> 00:14:08,660 in this whole class. 316 00:14:08,660 --> 00:14:12,710 And I know that it seemed like, oh, why are we teaching this? 317 00:14:12,710 --> 00:14:15,870 But it's this idea that you can take a whole bunch of data, 318 00:14:15,870 --> 00:14:17,810 put it into this thing called a hash function, 319 00:14:17,810 --> 00:14:19,340 and it commits to it. 320 00:14:19,340 --> 00:14:22,520 Because if you change any piece of the data-- 321 00:14:22,520 --> 00:14:24,700 any individual piece-- you're going 322 00:14:24,700 --> 00:14:28,380 to get a different output. 323 00:14:28,380 --> 00:14:32,928 And so it's really the key to it is that it commits something. 324 00:14:32,928 --> 00:14:35,220 You can do it with the New York Times crossword puzzle, 325 00:14:35,220 --> 00:14:41,490 and that was just my observation on it. 326 00:14:41,490 --> 00:14:43,380 So it's about data commitment. 327 00:14:43,380 --> 00:14:46,860 The other thing that Satoshi Nakamoto used 328 00:14:46,860 --> 00:14:51,750 was asymmetric cryptography, or digital signatures. 329 00:14:51,750 --> 00:14:54,660 I pulled one of the graphics that we talked about. 330 00:14:54,660 --> 00:14:59,850 But the idea being that, long ago, cryptography 331 00:14:59,850 --> 00:15:01,980 was, how do I encrypt something because I'm 332 00:15:01,980 --> 00:15:04,910 in a military campaign? 333 00:15:04,910 --> 00:15:07,520 And I would, maybe, just transpose. 334 00:15:07,520 --> 00:15:10,550 The simplest cryptography from thousands of years 335 00:15:10,550 --> 00:15:13,550 ago was transposing every letter. 336 00:15:13,550 --> 00:15:16,130 An A becomes a B, a B becomes a C. 337 00:15:16,130 --> 00:15:18,740 That would be a transposition of one. 338 00:15:18,740 --> 00:15:23,640 Well, that's pretty hard to imagine that anybody got away 339 00:15:23,640 --> 00:15:24,140 with that. 340 00:15:24,140 --> 00:15:26,990 But that was early cryptography, thousands of years ago-- just 341 00:15:26,990 --> 00:15:29,780 simple transposition. 342 00:15:29,780 --> 00:15:32,330 By the mid-20th century, there was 343 00:15:32,330 --> 00:15:35,150 very complex transpositions, which 344 00:15:35,150 --> 00:15:38,240 were called keys and enigma. 345 00:15:38,240 --> 00:15:41,900 If you ever go back and watch that movie Imitation Games, 346 00:15:41,900 --> 00:15:45,110 there was five metal rotors-- 347 00:15:45,110 --> 00:15:46,910 they were electronic rotors-- 348 00:15:46,910 --> 00:15:48,160 that changed. 349 00:15:48,160 --> 00:15:54,110 And every day, the Germans would change the five rotors. 350 00:15:54,110 --> 00:15:56,900 But still, they were symmetric cryptography. 351 00:15:56,900 --> 00:16:02,900 The U boats actually had the same five rotors 352 00:16:02,900 --> 00:16:06,440 as high command out of-- 353 00:16:06,440 --> 00:16:09,070 Berlin, let's say. 354 00:16:09,070 --> 00:16:11,510 So somebody came along in the 1970s 355 00:16:11,510 --> 00:16:13,220 and said, what if the key is different? 356 00:16:13,220 --> 00:16:14,210 It's asymmetric. 357 00:16:14,210 --> 00:16:16,140 There's a private key, and a public key. 358 00:16:16,140 --> 00:16:21,550 The public key, everybody gets, but the private key, you keep. 359 00:16:21,550 --> 00:16:23,800 And some of those computer sites-- 360 00:16:23,800 --> 00:16:26,680 cryptographers-- are here, right here, at MIT. 361 00:16:26,680 --> 00:16:30,730 Ron Rivest teaches a course in cryptography. 362 00:16:30,730 --> 00:16:34,440 But that invention, in the 1970s, is really key. 363 00:16:34,440 --> 00:16:36,520 So it's hash functions, and this concept 364 00:16:36,520 --> 00:16:39,110 of a private key and a public key, 365 00:16:39,110 --> 00:16:42,500 and of course a digital signature that goes with it. 366 00:16:42,500 --> 00:16:46,040 You probably don't even need to remember much about this class 367 00:16:46,040 --> 00:16:46,557 about this. 368 00:16:46,557 --> 00:16:48,140 But just know there's something called 369 00:16:48,140 --> 00:16:50,810 a private key and a public key. 370 00:16:50,810 --> 00:16:54,740 And then, Nakamoto's real innovation 371 00:16:54,740 --> 00:16:56,480 was off of Adam Back. 372 00:16:56,480 --> 00:17:01,340 And Adam Back said, if I use some computational work-- 373 00:17:01,340 --> 00:17:05,210 if I use computational work to find a hash function that 374 00:17:05,210 --> 00:17:08,930 has a predetermined number of zeros, 375 00:17:08,930 --> 00:17:10,310 it's not much of a puzzle. 376 00:17:10,310 --> 00:17:12,380 It's a random number generation. 377 00:17:12,380 --> 00:17:16,099 People will say proof-of-work is a computational puzzle. 378 00:17:16,099 --> 00:17:19,150 It is simply running-- 379 00:17:19,150 --> 00:17:21,770 this thing called mining is running-- electricity 380 00:17:21,770 --> 00:17:24,454 until you get the right number of leading zeros. 381 00:17:27,140 --> 00:17:29,270 But it can be efficiently verified. 382 00:17:29,270 --> 00:17:32,810 So the characteristics of proof-of-work is of work is-- 383 00:17:32,810 --> 00:17:34,760 invented in 1997. 384 00:17:34,760 --> 00:17:38,360 Blockchains were invented in 1991. 385 00:17:38,360 --> 00:17:40,700 But what Satoshi Nakamoto-- 386 00:17:40,700 --> 00:17:45,500 what she did was she pulled that together after like 20 or 30 387 00:17:45,500 --> 00:17:50,180 failed attempts to have a digital, non-central money 388 00:17:50,180 --> 00:17:53,930 and said, that that's where blockchain inspiration is. 389 00:17:56,470 --> 00:17:59,200 And so proof of work-- 390 00:17:59,200 --> 00:18:00,650 I use this little visual. 391 00:18:00,650 --> 00:18:06,430 But each block-- do you remember what the word nonce is? 392 00:18:06,430 --> 00:18:07,760 What's a nonce? 393 00:18:07,760 --> 00:18:10,350 Ross? 394 00:18:10,350 --> 00:18:13,000 Make me proud, somebody. 395 00:18:13,000 --> 00:18:13,665 Can't remember. 396 00:18:13,665 --> 00:18:14,467 AUDIENCE: And once. 397 00:18:14,467 --> 00:18:15,550 GARY GENSLER: What's that? 398 00:18:15,550 --> 00:18:16,930 AUDIENCE: And once. 399 00:18:16,930 --> 00:18:18,865 GARY GENSLER: And once-- a number once. 400 00:18:18,865 --> 00:18:19,990 AUDIENCE: Number used once. 401 00:18:19,990 --> 00:18:20,668 AUDIENCE: Once. 402 00:18:20,668 --> 00:18:21,960 GARY GENSLER: Number used once. 403 00:18:24,490 --> 00:18:28,070 And so why is a nonce important for proof of work? 404 00:18:28,070 --> 00:18:29,800 AUDIENCE: Because it's a unique number. 405 00:18:29,800 --> 00:18:31,450 GARY GENSLER: It's a unique number. 406 00:18:31,450 --> 00:18:34,810 AUDIENCE: And so establishing the integrity 407 00:18:34,810 --> 00:18:35,920 of that [INAUDIBLE]. 408 00:18:35,920 --> 00:18:38,200 GARY GENSLER: So all this whole proof of work-- 409 00:18:38,200 --> 00:18:40,930 I'm just going back as a review, but the whole proof of work 410 00:18:40,930 --> 00:18:43,930 is your computer is going to randomly keep 411 00:18:43,930 --> 00:18:47,080 picking one thing that changes. 412 00:18:47,080 --> 00:18:49,630 All the data, all the transactions are the same, 413 00:18:49,630 --> 00:18:53,730 but keep changing this thing called a nonce 414 00:18:53,730 --> 00:18:58,320 until I get some leading number of zeros. 415 00:18:58,320 --> 00:19:01,740 And the whole Bitcoin network around the globe-- 416 00:19:01,740 --> 00:19:04,860 all it's really doing is running a random number generation, 417 00:19:04,860 --> 00:19:09,180 or nonces, until the hash function 418 00:19:09,180 --> 00:19:13,600 is solved to have 18 leading zeros right now. 419 00:19:13,600 --> 00:19:14,950 Or I should say-- 420 00:19:14,950 --> 00:19:20,668 18 leading zeros when Bitcoin's value was $6,500 a Bitcoin. 421 00:19:20,668 --> 00:19:22,210 It wouldn't surprise me if it's going 422 00:19:22,210 --> 00:19:26,190 to fall down to 17 leading zeros, or 16 leading zeros, 423 00:19:26,190 --> 00:19:28,860 because that would make it easier to find. 424 00:19:31,770 --> 00:19:36,690 All of that cryptography is just to make the database structure. 425 00:19:36,690 --> 00:19:42,290 This database structure that is in blockchain-- 426 00:19:42,290 --> 00:19:43,560 to make this work. 427 00:19:43,560 --> 00:19:46,860 All that cryptography of hash functions, digital signatures, 428 00:19:46,860 --> 00:19:49,440 and a little proof of work is to make it 429 00:19:49,440 --> 00:19:53,010 what's called nearly immutable. 430 00:19:53,010 --> 00:19:55,590 And I'm putting the word "nearly" in because nothing's 431 00:19:55,590 --> 00:19:56,590 truly immutable. 432 00:19:56,590 --> 00:20:00,770 You could overwhelm even the Bitcoin network. 433 00:20:00,770 --> 00:20:04,745 Questions about the technology? 434 00:20:04,745 --> 00:20:05,430 [INAUDIBLE] 435 00:20:05,430 --> 00:20:07,930 AUDIENCE: Could you just go over the number of leading zeros 436 00:20:07,930 --> 00:20:09,510 concept again? 437 00:20:09,510 --> 00:20:15,800 GARY GENSLER: So the concept behind the proof of work 438 00:20:15,800 --> 00:20:25,560 that Adam Back came up with in the late 1990s was-- 439 00:20:25,560 --> 00:20:30,780 gotta run some data through a hash function. 440 00:20:30,780 --> 00:20:34,020 In his case, in Adam Back's case, it was emails. 441 00:20:34,020 --> 00:20:39,030 Literally run an email all the way through a hash function, 442 00:20:39,030 --> 00:20:42,810 and then add an additional bit of data to it-- 443 00:20:42,810 --> 00:20:46,930 this nonce, a number that's used once-- 444 00:20:46,930 --> 00:20:51,890 and run that nonce randomly until the hash that comes out. 445 00:20:51,890 --> 00:20:53,700 And my little visual at the bottom 446 00:20:53,700 --> 00:20:56,790 there has some leading zeros to it. 447 00:20:59,860 --> 00:21:01,360 AUDIENCE: So beginning of the hash 448 00:21:01,360 --> 00:21:03,610 will be its number of zeros? 449 00:21:03,610 --> 00:21:04,360 GARY GENSLER: Yes. 450 00:21:04,360 --> 00:21:07,350 So in this case, four leading zeros-- 451 00:21:07,350 --> 00:21:09,790 in that little visual there. 452 00:21:09,790 --> 00:21:12,280 All the data that goes into a block-- 453 00:21:12,280 --> 00:21:16,960 in Bitcoin, 1,000 to 1,500 transactions all go in. 454 00:21:16,960 --> 00:21:19,360 They're rolled up in this thing called a Merkle tree, 455 00:21:19,360 --> 00:21:21,700 but it all goes in. 456 00:21:21,700 --> 00:21:23,440 And the only thing that's changing 457 00:21:23,440 --> 00:21:28,580 is the miners are trying to find a random nonce 458 00:21:28,580 --> 00:21:33,810 until the hash has certain number of leading zeros. 459 00:21:33,810 --> 00:21:38,580 When Nakamoto wrote the paper, the difficulty was, I think, 460 00:21:38,580 --> 00:21:40,630 10 leading zeros. 461 00:21:40,630 --> 00:21:43,815 And you could solve that on a laptop. 462 00:21:47,360 --> 00:21:50,280 With eight more leading zeros-- 463 00:21:50,280 --> 00:21:52,320 and since it's hexadecimal-- 464 00:21:52,320 --> 00:21:56,700 I guess it's 16 to the eighth power 465 00:21:56,700 --> 00:22:04,477 harder, or something like seven trillion times harder. 466 00:22:04,477 --> 00:22:05,060 AUDIENCE: Yes. 467 00:22:10,460 --> 00:22:12,580 AUDIENCE: Let's assume that you and I were trying 468 00:22:12,580 --> 00:22:14,640 to have a transaction today. 469 00:22:14,640 --> 00:22:18,849 And I'm going to keep some coins. 470 00:22:18,849 --> 00:22:22,060 And I'm [INAUDIBLE] these for everyone 471 00:22:22,060 --> 00:22:24,120 who's transacting many, many times 472 00:22:24,120 --> 00:22:30,370 and adding layers of difficulty to the transaction that we had. 473 00:22:30,370 --> 00:22:35,210 If I want to use my balance in five years, 474 00:22:35,210 --> 00:22:37,800 how do we know what my balance is? 475 00:22:37,800 --> 00:22:41,000 I need to go back through all the history 476 00:22:41,000 --> 00:22:42,250 to know that I have a balance? 477 00:22:42,250 --> 00:22:47,350 GARY GENSLER: So at the heart of the blockchain technology, 478 00:22:47,350 --> 00:22:50,410 the Bitcoin network, but also all the other networks, 479 00:22:50,410 --> 00:22:54,130 is the entire transaction history is public. 480 00:22:54,130 --> 00:22:56,230 This both gives us its strength, but it's also 481 00:22:56,230 --> 00:22:58,870 one of its challenges. 482 00:22:58,870 --> 00:23:02,470 And so, yes, even five years from now, 483 00:23:02,470 --> 00:23:08,300 if you want to use one of your Bitcoins-- 484 00:23:08,300 --> 00:23:11,680 which are stored on the blockchain-- 485 00:23:11,680 --> 00:23:14,890 you have to be able to show that here is a transaction 486 00:23:14,890 --> 00:23:16,390 output from five years ago. 487 00:23:16,390 --> 00:23:21,710 If you remember, there is both transaction output models-- 488 00:23:21,710 --> 00:23:26,480 Bitcoin is a system that keeps all the individual transactions 489 00:23:26,480 --> 00:23:28,780 in a set called the UTXO-- 490 00:23:28,780 --> 00:23:32,920 the Unspent Transaction Output Set. 491 00:23:32,920 --> 00:23:37,300 You'd have to be able to show that you have that in 25 years, 492 00:23:37,300 --> 00:23:38,050 you said? 493 00:23:38,050 --> 00:23:44,200 2023-- in 2023, you'd have to be able to show that. 494 00:23:44,200 --> 00:23:46,810 AUDIENCE: So it's also a balance. 495 00:23:46,810 --> 00:23:51,260 They all spent this balance in the transaction [INAUDIBLE] 496 00:23:51,260 --> 00:23:53,060 a record. 497 00:23:53,060 --> 00:23:55,270 GARY GENSLER: It is a record of an output 498 00:23:55,270 --> 00:23:57,830 from an individual transaction. 499 00:23:57,830 --> 00:24:01,620 So yes, it is like a balance. 500 00:24:01,620 --> 00:24:07,610 But because Bitcoin does not add up all of those individual-- 501 00:24:07,610 --> 00:24:12,410 if you had 13 individual outputs, 502 00:24:12,410 --> 00:24:16,450 the Bitcoin network will not add up those 13. 503 00:24:16,450 --> 00:24:21,226 You'd have to use each of those individually as new inputs. 504 00:24:21,226 --> 00:24:26,170 AUDIENCE: And that one is on the last block, as well. 505 00:24:26,170 --> 00:24:28,627 But I'd have to go back [INAUDIBLE].. 506 00:24:28,627 --> 00:24:30,460 GARY GENSLER: All right, so the question is, 507 00:24:30,460 --> 00:24:31,600 does the last block-- 508 00:24:31,600 --> 00:24:36,190 Bitcoin has 550-some 1,000 blocks of data right now. 509 00:24:36,190 --> 00:24:40,040 Does the last block store all of the data? 510 00:24:40,040 --> 00:24:45,710 And the answer is, no, but it does store a hash function 511 00:24:45,710 --> 00:24:49,160 from the previous block, which stores a hash function 512 00:24:49,160 --> 00:24:51,100 from the previous block. 513 00:24:51,100 --> 00:24:57,450 So your transaction output might be on block number 420,000, 514 00:24:57,450 --> 00:25:01,760 130,000 blocks ago. 515 00:25:01,760 --> 00:25:04,430 But what shows that it's verifiable 516 00:25:04,430 --> 00:25:08,480 is that each one of these blocks are linked to another block 517 00:25:08,480 --> 00:25:10,460 through this thing called a commitment scheme 518 00:25:10,460 --> 00:25:12,790 called a hash function. 519 00:25:12,790 --> 00:25:17,120 So technically speaking, the latest block 520 00:25:17,120 --> 00:25:19,870 has a number in it-- 521 00:25:19,870 --> 00:25:22,750 a hexadecimal number, but a number in it-- 522 00:25:22,750 --> 00:25:25,660 called a hash function that is committed to. 523 00:25:25,660 --> 00:25:28,870 Because all the data that's come before it-- 524 00:25:28,870 --> 00:25:31,770 all the millions of transactions before it-- 525 00:25:31,770 --> 00:25:34,060 if somebody even changes one of them, 526 00:25:34,060 --> 00:25:37,672 you'll know that it's been tampered. 527 00:25:37,672 --> 00:25:41,460 AUDIENCE: Let's say I want to use that balance I have. 528 00:25:41,460 --> 00:25:45,500 Where do the blockchain verify that I have that? 529 00:25:45,500 --> 00:25:48,180 Is it on the last block, or does it 530 00:25:48,180 --> 00:25:53,495 go back as long as it finds the [INAUDIBLE]?? 531 00:25:53,495 --> 00:25:54,870 GARY GENSLER: In essence, it does 532 00:25:54,870 --> 00:25:59,420 both, because the last block better be valid. 533 00:25:59,420 --> 00:26:01,450 But for you to spend something, it 534 00:26:01,450 --> 00:26:10,180 has to really go back and find that the input that you now 535 00:26:10,180 --> 00:26:15,420 want to use is an output that has not yet been spent. 536 00:26:15,420 --> 00:26:18,570 Eric, did you have something that you were trying to-- 537 00:26:18,570 --> 00:26:21,660 AUDIENCE: Yeah, it was a very nice explanation. 538 00:26:21,660 --> 00:26:24,780 You have to go all the way where your transaction-- where 539 00:26:24,780 --> 00:26:29,365 you got the Bitcoin was, whatever 540 00:26:29,365 --> 00:26:32,225 block it is [INAUDIBLE] through the blockchain. 541 00:26:32,225 --> 00:26:35,470 If you would try to get those coins 542 00:26:35,470 --> 00:26:41,160 and use them as a [INAUDIBLE] for whatever transaction 543 00:26:41,160 --> 00:26:44,045 you want to do, you have to go all the way back. 544 00:26:44,045 --> 00:26:45,420 GARY GENSLER: And this complexity 545 00:26:45,420 --> 00:26:47,310 is part of the reason why it doesn't 546 00:26:47,310 --> 00:26:49,110 have the same performance-- 547 00:26:49,110 --> 00:26:53,130 doesn't have the same scalability-- yet. 548 00:26:53,130 --> 00:26:56,550 I'm one that believes in a-- 549 00:26:56,550 --> 00:26:59,100 I'm optimistic that some of these scalability 550 00:26:59,100 --> 00:27:01,230 and performance issues will be overcome. 551 00:27:01,230 --> 00:27:04,050 But it might take three, five, seven years. 552 00:27:04,050 --> 00:27:07,513 And it's not three or seven months. 553 00:27:07,513 --> 00:27:09,180 But you're absolutely right, because you 554 00:27:09,180 --> 00:27:12,210 have to kind of look back-- 555 00:27:12,210 --> 00:27:14,240 way back. 556 00:27:14,240 --> 00:27:18,860 The Ethereum network doesn't just do it on transactions. 557 00:27:18,860 --> 00:27:19,838 It has balances. 558 00:27:19,838 --> 00:27:21,880 And so it's a little bit more efficient that way, 559 00:27:21,880 --> 00:27:24,650 and has other inefficiencies-- 560 00:27:24,650 --> 00:27:30,450 many other big inefficiencies, but in some ways it-- 561 00:27:30,450 --> 00:27:32,100 We talked about smart contracts. 562 00:27:32,100 --> 00:27:36,360 Larry Lessig actually spoke to us also about smart contracts. 563 00:27:36,360 --> 00:27:38,970 Nick Szabo wrote about this in 1996. 564 00:27:38,970 --> 00:27:40,980 I like to talk about the history to say, 565 00:27:40,980 --> 00:27:44,670 it's not all about this one person that we don't know-- 566 00:27:44,670 --> 00:27:47,460 Satoshi Nakamoto-- but a set of promises 567 00:27:47,460 --> 00:27:50,550 specified in digital form, including protocols, where 568 00:27:50,550 --> 00:27:52,380 parties keep their promises. 569 00:27:52,380 --> 00:27:55,170 Think about it as a modern way to automate. 570 00:27:55,170 --> 00:27:57,375 So if you're out in business in the future, 571 00:27:57,375 --> 00:27:59,250 you're thinking about investing, and somebody 572 00:27:59,250 --> 00:28:02,700 raises a smart contract issue. 573 00:28:02,700 --> 00:28:05,447 Even if you're a Bitcoin minimalist, 574 00:28:05,447 --> 00:28:06,780 you might say, wait, wait, wait. 575 00:28:06,780 --> 00:28:09,000 There might be some way I can automate 576 00:28:09,000 --> 00:28:12,570 that which is currently done by lawyers, 577 00:28:12,570 --> 00:28:17,375 or by back-office folks, or accountants. 578 00:28:17,375 --> 00:28:18,750 So there's still something there. 579 00:28:18,750 --> 00:28:21,420 And we talked about a use case where the SDA-- 580 00:28:21,420 --> 00:28:25,260 the Swap Derivatives Association-- 581 00:28:25,260 --> 00:28:27,270 is looking at smart contracts to try 582 00:28:27,270 --> 00:28:34,970 to automate some of the previously written mechanisms. 583 00:28:34,970 --> 00:28:38,590 So smart contracts, which have come out 584 00:28:38,590 --> 00:28:43,480 of this whole blockchain technology movement, 585 00:28:43,480 --> 00:28:46,083 actually predates blockchain technology. 586 00:28:46,083 --> 00:28:48,250 But blockchain technology, again, as well-- maybe we 587 00:28:48,250 --> 00:28:53,780 can do this in a verifiable, nearly immutable way, as well. 588 00:28:53,780 --> 00:28:56,800 So moving value or automating some forms 589 00:28:56,800 --> 00:29:00,070 of contractual arrangement. 590 00:29:00,070 --> 00:29:01,950 But remember, they're not so smart, 591 00:29:01,950 --> 00:29:05,670 and they're not so much contracts. 592 00:29:05,670 --> 00:29:07,510 You probably still need the courts, 593 00:29:07,510 --> 00:29:09,490 and there's still some ambiguity that can't 594 00:29:09,490 --> 00:29:12,790 be pushed into these documents. 595 00:29:12,790 --> 00:29:14,710 So then, we talked a lot about the economics 596 00:29:14,710 --> 00:29:16,630 of blockchain technology. 597 00:29:16,630 --> 00:29:21,040 And it's around verification costs and networking costs. 598 00:29:21,040 --> 00:29:23,350 But it's about lowering verification costs. 599 00:29:23,350 --> 00:29:24,910 And some of those verification costs 600 00:29:24,910 --> 00:29:28,060 might just be the economic rents of the current incumbents. 601 00:29:28,060 --> 00:29:32,230 So blockchain technology may be a really important way 602 00:29:32,230 --> 00:29:37,630 to lower direct verification costs and deal with censorship. 603 00:29:37,630 --> 00:29:40,810 And I think that was a lot of what Nakamoto's innovation was 604 00:29:40,810 --> 00:29:42,850 about, was dealing with censorship that somebody 605 00:29:42,850 --> 00:29:45,400 could deny me the right to use my money when 606 00:29:45,400 --> 00:29:47,410 I want to use my money. 607 00:29:47,410 --> 00:29:49,810 But it could also be about lowering economic rents. 608 00:29:52,880 --> 00:29:54,142 Any questions? 609 00:29:57,450 --> 00:29:59,580 And then we said, how do you assess a use case? 610 00:29:59,580 --> 00:30:04,080 And this was at the core of even my CoinDesk article. 611 00:30:04,080 --> 00:30:07,500 Before you get to anything else is, how is it going to be, 612 00:30:07,500 --> 00:30:08,150 really? 613 00:30:08,150 --> 00:30:10,140 It's what's the value creation proposition? 614 00:30:10,140 --> 00:30:12,900 Again, what are the verification and networking costs that 615 00:30:12,900 --> 00:30:14,710 are really going to be reduced? 616 00:30:14,710 --> 00:30:16,650 And what are competitors doing? 617 00:30:16,650 --> 00:30:18,420 With a keen eye on-- 618 00:30:18,420 --> 00:30:21,810 in trade finance, there's 20 or 30 efforts, 619 00:30:21,810 --> 00:30:24,390 so it's worthwhile looking at those 20 or 30. 620 00:30:27,330 --> 00:30:30,360 I don't know which field you all go into, 621 00:30:30,360 --> 00:30:32,940 but if you're in a field where nobody has used blockchain 622 00:30:32,940 --> 00:30:35,910 technology, no competitor, then ask the question, 623 00:30:35,910 --> 00:30:40,290 well, how are competitors using traditional databases? 624 00:30:40,290 --> 00:30:43,340 And that might be the opportunity for you. 625 00:30:43,340 --> 00:30:45,350 And it might be a golden opportunity. 626 00:30:45,350 --> 00:30:48,660 But always look-- what's the value proposition? 627 00:30:48,660 --> 00:30:50,760 What verification and networking cost? 628 00:30:50,760 --> 00:30:52,260 What's the competition? 629 00:30:52,260 --> 00:30:58,273 And then, you get into the sort of tough details. 630 00:30:58,273 --> 00:30:59,690 And if you're investing, if you're 631 00:30:59,690 --> 00:31:01,400 a venture capitalist and somebody 632 00:31:01,400 --> 00:31:05,210 can't really tell you why multiple stakeholders need 633 00:31:05,210 --> 00:31:08,870 to write to the ledger, and what's their data-- 634 00:31:08,870 --> 00:31:11,780 what data are they going to be storing on this append-only 635 00:31:11,780 --> 00:31:13,380 log-- 636 00:31:13,380 --> 00:31:17,520 then I might suggest you invest in something else. 637 00:31:17,520 --> 00:31:19,270 I mean, you can always invest in something 638 00:31:19,270 --> 00:31:22,720 because you think that it's a momentum play. 639 00:31:22,720 --> 00:31:26,140 And if you invest, it's an early seed investment, 640 00:31:26,140 --> 00:31:28,108 and somebody will bail you out. 641 00:31:28,108 --> 00:31:29,650 But most venture capitalists-- you're 642 00:31:29,650 --> 00:31:33,380 investing for three to six years, or three to five years. 643 00:31:33,380 --> 00:31:35,350 And so I wouldn't always expect somebody 644 00:31:35,350 --> 00:31:37,220 is going to bail you out. 645 00:31:37,220 --> 00:31:38,590 Ask the tough questions. 646 00:31:38,590 --> 00:31:39,382 Jihee? 647 00:31:39,382 --> 00:31:41,590 AUDIENCE: But just because that is the best solution, 648 00:31:41,590 --> 00:31:45,220 it doesn't mean that they're going to be successful later, 649 00:31:45,220 --> 00:31:46,030 right? 650 00:31:46,030 --> 00:31:51,280 There are multiple examples of products, or services, 651 00:31:51,280 --> 00:31:53,830 or [INAUDIBLE] that are clearly not the best solution 652 00:31:53,830 --> 00:31:57,760 but have taken dominant position in a market. 653 00:31:57,760 --> 00:32:03,610 So I was curious, should it be the best solution? 654 00:32:03,610 --> 00:32:07,710 GARY GENSLER: Well, I think that's one of the questions. 655 00:32:07,710 --> 00:32:10,605 But you're saying, well, what if it's not the best solution? 656 00:32:14,340 --> 00:32:17,140 Going back to what Jeff Sprecher said, 657 00:32:17,140 --> 00:32:23,050 can it do something cheaper, faster, or better? 658 00:32:23,050 --> 00:32:25,468 And it might not be the best solution. 659 00:32:25,468 --> 00:32:27,260 But if it can at least answer the question, 660 00:32:27,260 --> 00:32:32,310 is it cheaper, better, or faster, 661 00:32:32,310 --> 00:32:34,380 you're probably not going to be successful. 662 00:32:34,380 --> 00:32:37,070 And in some cases, it's not the best data solution, 663 00:32:37,070 --> 00:32:40,380 but there's such high economic rents 664 00:32:40,380 --> 00:32:42,980 that however the mortgage product, the health care 665 00:32:42,980 --> 00:32:45,230 product, the product that is being delivered now, 666 00:32:45,230 --> 00:32:49,080 the payment solution that you're saying, ah, 667 00:32:49,080 --> 00:32:50,290 it's not the best solution. 668 00:32:50,290 --> 00:32:53,910 But this is my solution on how I'm going to do it cheaper, 669 00:32:53,910 --> 00:32:57,000 because they have so many economic rents. 670 00:32:57,000 --> 00:32:59,760 So maybe that the word should say, 671 00:32:59,760 --> 00:33:03,900 why is this a better solution, rather than best solution. 672 00:33:03,900 --> 00:33:06,294 I'll agree with that. 673 00:33:06,294 --> 00:33:07,890 Other queries? 674 00:33:10,490 --> 00:33:13,550 And then, if there's a native token, 675 00:33:13,550 --> 00:33:16,490 I think that's where you start to lose May. 676 00:33:16,490 --> 00:33:19,850 But now, I'm going to ask the class. 677 00:33:19,850 --> 00:33:23,030 How many people have come out of this semester thinking 678 00:33:23,030 --> 00:33:26,450 there's a lot of initial coin offerings 679 00:33:26,450 --> 00:33:29,978 that you'd want to personally invest in? 680 00:33:29,978 --> 00:33:31,320 [LAUGHTER] 681 00:33:31,320 --> 00:33:33,590 All right, is there anybody in the class 682 00:33:33,590 --> 00:33:38,340 that owns an initial coin offering-- a non-Bitcoin token? 683 00:33:38,340 --> 00:33:42,741 OK, do you want to tell us why? 684 00:33:42,741 --> 00:33:44,330 AUDIENCE: Yeah. 685 00:33:44,330 --> 00:33:48,830 GARY GENSLER: Give us the value proposition, or the pitch. 686 00:33:48,830 --> 00:33:52,710 AUDIENCE: So I went to work for a startup, [INAUDIBLE] 687 00:33:52,710 --> 00:33:53,780 this summer. 688 00:33:53,780 --> 00:33:57,050 And after the summer, they paid me US dollars. 689 00:33:57,050 --> 00:34:01,930 But after the summer, they gave me a lot of tokens they have. 690 00:34:01,930 --> 00:34:03,520 And I saved a little bit of them. 691 00:34:03,520 --> 00:34:04,520 GARY GENSLER: All right. 692 00:34:04,520 --> 00:34:06,380 So you got paid in tokens. 693 00:34:06,380 --> 00:34:09,030 AUDIENCE: No, I was paid in US dollars. 694 00:34:09,030 --> 00:34:12,400 After the summer, they gave me most of the-- 695 00:34:12,400 --> 00:34:16,030 like a bunch of-- some of the tokens. 696 00:34:16,030 --> 00:34:17,409 GARY GENSLER: As a bonus or? 697 00:34:17,409 --> 00:34:18,040 AUDIENCE: As a bonus, yeah. 698 00:34:18,040 --> 00:34:18,360 GARY GENSLER: I see. 699 00:34:18,360 --> 00:34:19,659 As a bonus, you got-- 700 00:34:19,659 --> 00:34:20,170 all right. 701 00:34:20,170 --> 00:34:21,610 And you sold some of them? 702 00:34:21,610 --> 00:34:23,260 AUDIENCE: Some of them, sold them. 703 00:34:23,260 --> 00:34:24,429 GARY GENSLER: And I see. 704 00:34:24,429 --> 00:34:26,270 AUDIENCE: Some of them I still [INAUDIBLE].. 705 00:34:26,270 --> 00:34:27,270 GARY GENSLER: All right. 706 00:34:27,270 --> 00:34:32,300 And do you think that there's a use for these native tokens? 707 00:34:32,300 --> 00:34:33,726 AUDIENCE: This specific startup? 708 00:34:33,726 --> 00:34:34,704 GARY GENSLER: Yeah. 709 00:34:34,704 --> 00:34:35,980 Names will not be repeated. 710 00:34:35,980 --> 00:34:36,912 It's being filmed. 711 00:34:36,912 --> 00:34:37,870 AUDIENCE: I don't know. 712 00:34:37,870 --> 00:34:38,995 GARY GENSLER: I don't know. 713 00:34:38,995 --> 00:34:39,760 Sean. 714 00:34:39,760 --> 00:34:41,739 AUDIENCE: So I thought of it-- well, 715 00:34:41,739 --> 00:34:44,520 I just thought a little of like Tron, which is a-- 716 00:34:44,520 --> 00:34:46,159 GARY GENSLER: Tron. 717 00:34:46,159 --> 00:34:47,409 AUDIENCE: Yeah, it's a gaming. 718 00:34:47,409 --> 00:34:48,820 GARY GENSLER: Gaming site. 719 00:34:48,820 --> 00:34:50,850 AUDIENCE: Which is a-- basically a company 720 00:34:50,850 --> 00:34:52,460 started by a friend of mine. 721 00:34:52,460 --> 00:34:53,980 So it's kind of a token of support 722 00:34:53,980 --> 00:34:55,219 instead of like [INAUDIBLE]. 723 00:34:55,219 --> 00:34:56,219 GARY GENSLER: All right. 724 00:34:56,219 --> 00:34:59,020 So you bought some tokens as a token of support 725 00:34:59,020 --> 00:35:01,630 for a friend, who started a company. 726 00:35:01,630 --> 00:35:04,330 Do you think they have some use? 727 00:35:04,330 --> 00:35:05,740 Can you use the token? 728 00:35:05,740 --> 00:35:07,778 AUDIENCE: I don't think there's-- well-- 729 00:35:07,778 --> 00:35:09,260 [LAUGHTER] 730 00:35:09,260 --> 00:35:11,800 Well, it's supposed to be the-- 731 00:35:11,800 --> 00:35:14,885 building up to the biggest gaming network [INAUDIBLE].. 732 00:35:14,885 --> 00:35:16,510 GARY GENSLER: But is it pre-functional, 733 00:35:16,510 --> 00:35:18,205 or is it functional? 734 00:35:18,205 --> 00:35:19,790 It's pre-functional. 735 00:35:19,790 --> 00:35:20,680 Hugo? 736 00:35:20,680 --> 00:35:21,575 AUDIENCE: Yeah. 737 00:35:21,575 --> 00:35:22,950 GARY GENSLER: You have-- you are? 738 00:35:22,950 --> 00:35:24,330 AUDIENCE: I have-- yeah. 739 00:35:24,330 --> 00:35:25,330 GARY GENSLER: All right. 740 00:35:25,330 --> 00:35:26,247 You don't-- all right. 741 00:35:26,247 --> 00:35:27,520 He's got a whole portfolio. 742 00:35:27,520 --> 00:35:28,842 I can tell. 743 00:35:28,842 --> 00:35:29,800 AUDIENCE: I don't know. 744 00:35:29,800 --> 00:35:32,862 I've been interested in this for over a year now. 745 00:35:32,862 --> 00:35:34,320 So I wanted to make a whole radical 746 00:35:34,320 --> 00:35:36,612 of like, oh, this one looks like it's going to go well, 747 00:35:36,612 --> 00:35:37,680 but this one's a scam. 748 00:35:37,680 --> 00:35:44,810 And there are some that seem like sure-- in the long run, 749 00:35:44,810 --> 00:35:47,760 they may retain some value. 750 00:35:47,760 --> 00:35:49,967 But I probably made some bad investments. 751 00:35:49,967 --> 00:35:50,800 GARY GENSLER: Right. 752 00:35:50,800 --> 00:35:54,340 So I don't want to make you sort of grieve those bad investment 753 00:35:54,340 --> 00:35:57,820 decisions, but are any of your tokens that you have held 754 00:35:57,820 --> 00:36:00,990 or hold currently-- are any of them functional? 755 00:36:00,990 --> 00:36:03,530 Like, they can be used on a network? 756 00:36:03,530 --> 00:36:04,200 AUDIENCE: Yeah. 757 00:36:04,200 --> 00:36:07,405 And I actually use at least one of them. 758 00:36:07,405 --> 00:36:08,830 GARY GENSLER: All right. 759 00:36:08,830 --> 00:36:10,758 Here. 760 00:36:10,758 --> 00:36:12,550 Are you an owner, or you've got a question? 761 00:36:12,550 --> 00:36:14,020 AUDIENCE: I'm an owner, but not-- 762 00:36:14,020 --> 00:36:15,510 I haven't voted. 763 00:36:15,510 --> 00:36:17,410 Both the tokens-- I'm mining the tokens. 764 00:36:17,410 --> 00:36:22,430 But in any case, I wouldn't disregard maybe 765 00:36:22,430 --> 00:36:27,860 tokens that's a valid way of generating a means to jumpstart 766 00:36:27,860 --> 00:36:34,840 a network and create some sort of valid way 767 00:36:34,840 --> 00:36:41,270 to drive collective action for some disintermediation effort, 768 00:36:41,270 --> 00:36:43,230 because I mean I-- 769 00:36:43,230 --> 00:36:45,820 and I think it's a little bit unfair 770 00:36:45,820 --> 00:36:50,920 to ask if we kind of join or support ICOs 771 00:36:50,920 --> 00:36:54,397 in the context of there's huge amount of scams 772 00:36:54,397 --> 00:36:56,230 that are [INAUDIBLE] right now in the media. 773 00:36:56,230 --> 00:37:00,037 But I wouldn't disregard the whole concept of tokens here. 774 00:37:00,037 --> 00:37:01,120 GARY GENSLER: [INAUDIBLE]. 775 00:37:01,120 --> 00:37:03,370 I admit it might have been an unfair question. 776 00:37:03,370 --> 00:37:04,720 I was just trying to get-- 777 00:37:04,720 --> 00:37:09,740 I was just trying to get some support in the room for tokens. 778 00:37:09,740 --> 00:37:12,550 AUDIENCE: The concept for me is absolutely valid. 779 00:37:12,550 --> 00:37:15,310 I would explore, seriously explore-- 780 00:37:15,310 --> 00:37:18,850 if I launch a startup in the blockchain space, 781 00:37:18,850 --> 00:37:20,890 seriously explore the possibility 782 00:37:20,890 --> 00:37:25,030 of using maybe tokens in the permissionless environment. 783 00:37:25,030 --> 00:37:26,500 So I suppose the permission-- 784 00:37:26,500 --> 00:37:31,720 GARY GENSLER: So I think that there is definitely 785 00:37:31,720 --> 00:37:37,830 a future in this overall technology of making 786 00:37:37,830 --> 00:37:42,820 a database harder to tamper with, 787 00:37:42,820 --> 00:37:45,460 more immutable on a scale, maybe not 788 00:37:45,460 --> 00:37:48,830 100% immutable, but closer to that immutable 789 00:37:48,830 --> 00:37:51,580 and inverifiable. 790 00:37:51,580 --> 00:37:56,060 I think that native tokens could help jump start a network. 791 00:37:56,060 --> 00:37:59,740 What I think the challenge is I think the challenge is, 792 00:37:59,740 --> 00:38:02,490 how do you compete with fiat currency 793 00:38:02,490 --> 00:38:07,110 that has such extraordinary network effects? 794 00:38:07,110 --> 00:38:10,790 So the Ernst & Young study, the recent study, I think 795 00:38:10,790 --> 00:38:14,870 captures this a bit when they looked at the 140 largest ICOs 796 00:38:14,870 --> 00:38:17,030 from 2017. 797 00:38:17,030 --> 00:38:20,390 Only about I think it was 17 of 140 798 00:38:20,390 --> 00:38:22,850 have a functioning network right now, 799 00:38:22,850 --> 00:38:26,330 where you can use the token, 13%. 800 00:38:26,330 --> 00:38:30,110 But of those 17, I think seven of them, 801 00:38:30,110 --> 00:38:32,360 the networks that are live, instead of, well, 802 00:38:32,360 --> 00:38:35,720 guess what, we'll also take fiat currency for this service 803 00:38:35,720 --> 00:38:37,850 that we're providing, whether they're providing 804 00:38:37,850 --> 00:38:39,600 file storage or something else. 805 00:38:39,600 --> 00:38:41,570 So they definitely used a native token 806 00:38:41,570 --> 00:38:45,050 to jumpstart the network as a crowdfunding, 807 00:38:45,050 --> 00:38:49,550 but then when they got to the later functional stage, 808 00:38:49,550 --> 00:38:51,710 the operator, the entrepreneur says, 809 00:38:51,710 --> 00:38:53,740 I don't want to limit my users. 810 00:38:53,740 --> 00:38:57,260 I'm going to also take fiat currency to buy or secure 811 00:38:57,260 --> 00:38:59,720 whatever service I'm having. 812 00:38:59,720 --> 00:39:02,840 And I think that's the business model challenge. 813 00:39:05,520 --> 00:39:10,110 And it may well be that there are networks, 814 00:39:10,110 --> 00:39:13,890 just like in gaming sites, where you want skins or shields, 815 00:39:13,890 --> 00:39:17,430 and there's something about only trading those tokens. 816 00:39:20,580 --> 00:39:23,250 But to date, that's been a limited class. 817 00:39:23,250 --> 00:39:28,680 That's where I probably will declare myself. 818 00:39:28,680 --> 00:39:31,860 I think, of course, there's a bunch of tradeoffs 819 00:39:31,860 --> 00:39:34,110 over the next three to seven years 820 00:39:34,110 --> 00:39:38,350 that will be sorted out around performance, privacy, security. 821 00:39:38,350 --> 00:39:39,970 We talked about layer 2. 822 00:39:39,970 --> 00:39:43,230 If any of you are actually interested in work 823 00:39:43,230 --> 00:39:45,090 in the blockchain technology field, 824 00:39:45,090 --> 00:39:47,670 you're going to learn a lot about layer 2 solutions 825 00:39:47,670 --> 00:39:51,630 and side chains and all the performance scalability things 826 00:39:51,630 --> 00:39:54,420 that colleagues over at the Digital Currency Initiative 827 00:39:54,420 --> 00:39:59,010 are spending a lot of time on this question right here. 828 00:39:59,010 --> 00:40:01,620 And then, of course, this is a business model. 829 00:40:01,620 --> 00:40:03,502 Just purely as a venture capitalist, 830 00:40:03,502 --> 00:40:05,460 you have to think about, how do I get adoption? 831 00:40:05,460 --> 00:40:07,770 How do-- what's my customer user interface? 832 00:40:07,770 --> 00:40:10,920 I'm sure, Iman, you're thinking a lot about this, right? 833 00:40:10,920 --> 00:40:13,860 You could have the best idea on a data structure. 834 00:40:13,860 --> 00:40:17,460 You could have a really nifty idea on a native token. 835 00:40:17,460 --> 00:40:19,590 But you still have to deal with whatever 836 00:40:19,590 --> 00:40:24,270 venture capitalist has to do is, how do I get users to use this? 837 00:40:24,270 --> 00:40:24,900 Adoption. 838 00:40:24,900 --> 00:40:27,120 What's my UI? 839 00:40:27,120 --> 00:40:32,310 And it's sort of a much more generic set of issues. 840 00:40:32,310 --> 00:40:32,810 Iman. 841 00:40:32,810 --> 00:40:34,980 AUDIENCE: That was my answer to Jihee. 842 00:40:34,980 --> 00:40:36,900 She's totally right. 843 00:40:36,900 --> 00:40:43,380 It's all about-- it's a tension between adoption and quality 844 00:40:43,380 --> 00:40:44,640 of product. 845 00:40:44,640 --> 00:40:48,188 And that's why we moved from Ethereum 846 00:40:48,188 --> 00:40:49,230 to the public blockchain. 847 00:40:49,230 --> 00:40:53,710 But we didn't think that it will be mass adopt by-- 848 00:40:53,710 --> 00:40:56,550 massively adopted by the incumbents. 849 00:40:56,550 --> 00:40:59,580 So we moved to a permission blockchain 850 00:40:59,580 --> 00:41:01,585 because of the adoption. 851 00:41:01,585 --> 00:41:03,960 GARY GENSLER: And I can't remember if you're able to say. 852 00:41:03,960 --> 00:41:05,430 Which are you using-- 853 00:41:05,430 --> 00:41:08,040 Hyperledger Fabric or Corda or? 854 00:41:08,040 --> 00:41:10,750 AUDIENCE: We're using Corda because of the state channels 855 00:41:10,750 --> 00:41:14,110 solution that they provide. 856 00:41:14,110 --> 00:41:17,260 GARY GENSLER: So we talked about frameworks. 857 00:41:17,260 --> 00:41:21,430 Over time, it's my thought that that orange line, the slope 858 00:41:21,430 --> 00:41:23,560 of that will come down, and we'll move more 859 00:41:23,560 --> 00:41:25,900 towards the decentralized side. 860 00:41:25,900 --> 00:41:30,100 But for the orange line, the slope to come down, in essence, 861 00:41:30,100 --> 00:41:34,540 the costs of the scalability, security, even the coordination 862 00:41:34,540 --> 00:41:35,740 needs to come down. 863 00:41:35,740 --> 00:41:38,920 It's always going to be some challenge on governance. 864 00:41:38,920 --> 00:41:42,490 In essence, who is going to pay for the software 865 00:41:42,490 --> 00:41:47,170 development in a truly decentralized network? 866 00:41:47,170 --> 00:41:52,900 And Bitcoin has been supported, for instance, by, really, 867 00:41:52,900 --> 00:41:56,200 an incredible group of Bitcoin Core developers, 868 00:41:56,200 --> 00:41:58,390 in part because they just believe 869 00:41:58,390 --> 00:42:00,520 in it, in part that some are funded 870 00:42:00,520 --> 00:42:04,530 by institutions like MIT. 871 00:42:04,530 --> 00:42:07,720 But it's still-- it's always a challenge in a truly 872 00:42:07,720 --> 00:42:09,310 decentralized space. 873 00:42:09,310 --> 00:42:10,570 Who's going to fund-- 874 00:42:10,570 --> 00:42:14,680 who's got the economic interest to fund the continued updates 875 00:42:14,680 --> 00:42:15,850 and software? 876 00:42:15,850 --> 00:42:21,280 I think Uber could be completely a decentralized network. 877 00:42:21,280 --> 00:42:25,210 And probably, drivers would get paid more. 878 00:42:25,210 --> 00:42:27,490 But then where would be the incentive? 879 00:42:27,490 --> 00:42:29,710 Who would work on the software development 880 00:42:29,710 --> 00:42:33,420 to update their ride app? 881 00:42:33,420 --> 00:42:37,310 So there's a key tradeoff there. 882 00:42:37,310 --> 00:42:40,340 The financial sector-- well, it moves and allocates 883 00:42:40,340 --> 00:42:41,510 money and risk. 884 00:42:41,510 --> 00:42:42,890 We talked about that. 885 00:42:42,890 --> 00:42:45,170 But it relies on a system of ledgers, 886 00:42:45,170 --> 00:42:50,390 and it's always had a symbiotic relationship with technology. 887 00:42:50,390 --> 00:42:53,450 The opportunities-- these are the key opportunities, 888 00:42:53,450 --> 00:42:55,460 which I think any thought fintech-- 889 00:42:55,460 --> 00:42:59,060 it's not just blockchain but even AI and machine learning-- 890 00:42:59,060 --> 00:43:01,370 has around finances. 891 00:43:01,370 --> 00:43:02,900 The customer interfaces-- there's 892 00:43:02,900 --> 00:43:04,850 a lot of legacy customer interfaces 893 00:43:04,850 --> 00:43:06,410 that you can actually compete with, 894 00:43:06,410 --> 00:43:08,150 if you're doing a startup. 895 00:43:08,150 --> 00:43:10,750 The economic rents are high. 896 00:43:10,750 --> 00:43:13,730 There's 7 and 1/2 percent of the economy is finance. 897 00:43:13,730 --> 00:43:16,340 Not all of the 7 and 1/2 percent is rents, 898 00:43:16,340 --> 00:43:19,470 but there is a lot of extra-- 899 00:43:19,470 --> 00:43:22,220 excuse the expression-- a lot of extra vig or juice 900 00:43:22,220 --> 00:43:26,330 in that model. 901 00:43:26,330 --> 00:43:29,300 And, of course, we've had repeated crises 902 00:43:29,300 --> 00:43:30,260 and instability. 903 00:43:30,260 --> 00:43:33,800 And financial inclusion-- we talked about products 904 00:43:33,800 --> 00:43:36,980 like Alipay and M-Pesa that were really 905 00:43:36,980 --> 00:43:41,430 about financial inclusion using new technologies. 906 00:43:41,430 --> 00:43:43,940 So what are the technologies in our time? 907 00:43:43,940 --> 00:43:47,840 I just put up a nice little graphic here. 908 00:43:47,840 --> 00:43:49,790 I'm doing some work with a group called 909 00:43:49,790 --> 00:43:53,810 FinTech at CSAIL, the Computer Science and AI Lab here. 910 00:43:53,810 --> 00:43:56,180 And when we leave with the funding companies, 911 00:43:56,180 --> 00:44:00,780 the big funding companies of that unit over there, 912 00:44:00,780 --> 00:44:03,635 blockchain is on their list, but it's not number one. 913 00:44:03,635 --> 00:44:06,625 To just put it in context, in terms of their technologies, 914 00:44:06,625 --> 00:44:08,250 they're thinking about machine learning 915 00:44:08,250 --> 00:44:12,270 and AI would probably be the biggest piece. 916 00:44:12,270 --> 00:44:15,300 But just to give a flavor for-- 917 00:44:15,300 --> 00:44:18,990 but it's one of the top three or four. 918 00:44:18,990 --> 00:44:22,050 But as we've talked about, they have challenges. 919 00:44:22,050 --> 00:44:25,040 The financial sector sees a lot of challenges. 920 00:44:25,040 --> 00:44:27,540 AUDIENCE: I just want to go back to previous slides, 921 00:44:27,540 --> 00:44:30,930 and how would the virtual reality impact finance? 922 00:44:30,930 --> 00:44:32,660 I'm not sure. 923 00:44:32,660 --> 00:44:33,160 [INAUDIBLE] 924 00:44:33,160 --> 00:44:34,535 GARY GENSLER: That has happened-- 925 00:44:34,535 --> 00:44:36,650 AUDIENCE: AR and VR, [INAUDIBLE] tech, finance. 926 00:44:41,150 --> 00:44:43,150 GARY GENSLER: I'm not familiar with the projects 927 00:44:43,150 --> 00:44:44,890 in that space. 928 00:44:44,890 --> 00:44:48,403 But where-- and I'm sure I'm getting now, 929 00:44:48,403 --> 00:44:50,320 you're going to inspire me to do some research 930 00:44:50,320 --> 00:44:53,770 in the next day or two to shoot your email with some projects. 931 00:44:53,770 --> 00:45:01,090 But virtual reality and hearing it does 932 00:45:01,090 --> 00:45:03,340 allow for more financial inclusion. 933 00:45:03,340 --> 00:45:07,750 And the question really is, can you just, rather-- 934 00:45:07,750 --> 00:45:09,460 think of the user experience. 935 00:45:09,460 --> 00:45:17,530 If you can find an enjoyable way to spend more money 936 00:45:17,530 --> 00:45:20,530 through virtual reality, the banking system 937 00:45:20,530 --> 00:45:22,900 wants to tap into that. 938 00:45:22,900 --> 00:45:27,010 It's not one of the top things at FinTech at CSAIL, 939 00:45:27,010 --> 00:45:31,060 I'd say AI, machine learning, biometrics, and blockchain. 940 00:45:31,060 --> 00:45:32,800 Everything's in the cloud, so they're not 941 00:45:32,800 --> 00:45:37,030 asking, right now, MIT to help them with cloud issues. 942 00:45:37,030 --> 00:45:40,120 Open API is a really big issue, particularly in London 943 00:45:40,120 --> 00:45:43,900 and in the UK, where it's been regulated that the banks need 944 00:45:43,900 --> 00:45:45,910 to use Open API. 945 00:45:45,910 --> 00:45:49,090 So I'd say AR and virtual reality 946 00:45:49,090 --> 00:45:53,110 is probably eighth on the list there. 947 00:45:53,110 --> 00:45:56,050 And I just include it, because I kind of think it's-- 948 00:45:56,050 --> 00:46:00,190 they'll find a way to change a user interface. 949 00:46:00,190 --> 00:46:02,950 We talked about there's a bunch of issues, 950 00:46:02,950 --> 00:46:05,830 and that's really leading to the use. 951 00:46:05,830 --> 00:46:08,560 We talked about traditional databases, which-- 952 00:46:08,560 --> 00:46:12,220 where parties can just create, read, write, 953 00:46:12,220 --> 00:46:14,650 the ledger, the private blockchain, which, 954 00:46:14,650 --> 00:46:21,220 of course, is what most of trade finance, most of payments, 955 00:46:21,220 --> 00:46:23,470 all the big central banks that have looked at it 956 00:46:23,470 --> 00:46:27,300 or moved from Ethereum to doing Hyperledger projects-- 957 00:46:27,300 --> 00:46:30,490 Singapore and Canada's Jasper-- 958 00:46:30,490 --> 00:46:32,650 and maybe permissionless. 959 00:46:32,650 --> 00:46:39,300 But the finance is really in the second bucket right now. 960 00:46:39,300 --> 00:46:42,480 Eric, it's not to say that native tokens won't take off 961 00:46:42,480 --> 00:46:46,300 and, Hugo, that your portfolio will be worth something, maybe. 962 00:46:49,240 --> 00:46:54,550 So here, we went through a bunch of use cases. 963 00:46:54,550 --> 00:46:57,750 But the biggest use case so far has been the $20 to $30 billion 964 00:46:57,750 --> 00:47:01,830 that's been raised through ICOs. 965 00:47:01,830 --> 00:47:06,950 I personally think that's going to come down, way down. 966 00:47:06,950 --> 00:47:09,410 But payment systems-- 967 00:47:09,410 --> 00:47:14,700 I mean we might differ on the views of XRP, a token, 968 00:47:14,700 --> 00:47:16,680 but Ripple and other companies have 969 00:47:16,680 --> 00:47:20,220 done a lot to try to be a catalyst for change and payment 970 00:47:20,220 --> 00:47:21,250 systems. 971 00:47:21,250 --> 00:47:22,950 We talked a lot about trade finance. 972 00:47:22,950 --> 00:47:24,720 Again, they're more in the permissions 973 00:47:24,720 --> 00:47:25,980 than the permissionless. 974 00:47:29,190 --> 00:47:31,990 Clearing and settling-- what's going on in Australia. 975 00:47:31,990 --> 00:47:35,970 Again, permissionless, not-- I mean permissioned, not 976 00:47:35,970 --> 00:47:38,910 permissionless. 977 00:47:38,910 --> 00:47:43,240 And then some of the non-financial-- 978 00:47:43,240 --> 00:47:46,290 had a lively discussion about supply chain management, 979 00:47:46,290 --> 00:47:51,160 and James, I'll forever remember that you're not a fan of it. 980 00:47:51,160 --> 00:47:55,330 But Walmart and Cargo might disagree with James. 981 00:47:55,330 --> 00:47:58,480 And they have active projects on it. 982 00:47:58,480 --> 00:48:01,160 And Lauren gave-- where's Lauren? 983 00:48:01,160 --> 00:48:03,790 She's here, maybe-- gave a reason why 984 00:48:03,790 --> 00:48:06,490 in supply chain sustainability that you 985 00:48:06,490 --> 00:48:08,080 might be on the other side. 986 00:48:08,080 --> 00:48:14,180 Digital ID was the last thing we talked about. 987 00:48:14,180 --> 00:48:17,890 So there's a lot going on, and when there's $20 or $30 billion 988 00:48:17,890 --> 00:48:19,930 raised, there's going to be-- these projects 989 00:48:19,930 --> 00:48:21,620 will have some life to them. 990 00:48:21,620 --> 00:48:26,410 We'll learn a lot from them along the way. 991 00:48:26,410 --> 00:48:28,720 Crypto finance-- when I first put this slide up, 992 00:48:28,720 --> 00:48:30,350 it wasn't $110 billion. 993 00:48:30,350 --> 00:48:32,590 And gosh knows I did this yesterday, 994 00:48:32,590 --> 00:48:35,350 so I don't even know what it is today. 995 00:48:35,350 --> 00:48:36,790 Pretty volatile market. 996 00:48:36,790 --> 00:48:41,480 Interestingly, Bitcoin staggered about 55% or 57% of the market. 997 00:48:41,480 --> 00:48:48,450 So there's a high correlation, though Ethereum slid, and XRP 998 00:48:48,450 --> 00:48:54,510 is number two in that market since we first got together 999 00:48:54,510 --> 00:48:56,640 in September. 1000 00:48:56,640 --> 00:49:01,290 Here, what I think the challenge is, as an investor-- 1001 00:49:01,290 --> 00:49:03,840 I'm not trying to pick on you, Hugo, but you know, 1002 00:49:03,840 --> 00:49:05,160 an investor. 1003 00:49:05,160 --> 00:49:07,770 But I think that's the number one challenge. 1004 00:49:07,770 --> 00:49:09,500 I really do. 1005 00:49:09,500 --> 00:49:11,430 What's the viability of that token? 1006 00:49:11,430 --> 00:49:12,680 And Eric, I'll agree with you. 1007 00:49:12,680 --> 00:49:15,120 Some tokens will be worth something. 1008 00:49:15,120 --> 00:49:16,700 But it's assessing that. 1009 00:49:16,700 --> 00:49:19,940 But the markets are readily susceptible to fraud. 1010 00:49:19,940 --> 00:49:22,160 How do I keep custody for the private key? 1011 00:49:22,160 --> 00:49:23,880 Fidelity has announced they're going 1012 00:49:23,880 --> 00:49:26,000 to have a new private key solution, 1013 00:49:26,000 --> 00:49:28,940 so it's interesting that big incumbents are coming in. 1014 00:49:28,940 --> 00:49:34,490 It's not just the Zappos and the early stage custody solutions. 1015 00:49:34,490 --> 00:49:38,330 Even Backed, the Intercontinental Exchange, 1016 00:49:38,330 --> 00:49:39,830 has put a lot of money into it. 1017 00:49:39,830 --> 00:49:42,380 And they'll, in essence, have a custody solution, 1018 00:49:42,380 --> 00:49:45,590 because you could buy Bitcoin, one day futures, 1019 00:49:45,590 --> 00:49:47,960 have a Bitcoin, and then they would, in essence, 1020 00:49:47,960 --> 00:49:49,210 hold custody for you. 1021 00:49:51,970 --> 00:49:54,950 And then we talked about crypto exchanges. 1022 00:49:54,950 --> 00:49:57,760 And the key thing-- and if you take anything away 1023 00:49:57,760 --> 00:50:00,460 about exchanges, other than right now they're unregulated 1024 00:50:00,460 --> 00:50:03,520 and they're probably highly susceptible to front running 1025 00:50:03,520 --> 00:50:06,880 and manipulation, is that they're a little bit different, 1026 00:50:06,880 --> 00:50:09,130 because they're not just matching 1027 00:50:09,130 --> 00:50:12,370 agents, which the London Stock Exchange or the New York Stock 1028 00:50:12,370 --> 00:50:15,430 Exchange are [INAUDIBLE],, but they also are counterparties 1029 00:50:15,430 --> 00:50:16,510 and custodians. 1030 00:50:19,330 --> 00:50:22,750 The other thing to take away about crypto exchanges 1031 00:50:22,750 --> 00:50:24,430 is they're highly centralized. 1032 00:50:24,430 --> 00:50:28,330 So it's a central irony that Satoshi Nakamoto is trying 1033 00:50:28,330 --> 00:50:32,190 to have decentralization, and yet there's 1034 00:50:32,190 --> 00:50:35,090 an awful lot of centralization. 1035 00:50:35,090 --> 00:50:37,580 So I'm just thinking about core takeaways. 1036 00:50:37,580 --> 00:50:39,980 If you're ever investing, or you're thinking about this, 1037 00:50:39,980 --> 00:50:42,560 or you're just involved in a dinner party conversation, 1038 00:50:42,560 --> 00:50:45,620 that central irony in a decentralized space, 1039 00:50:45,620 --> 00:50:51,530 95% of the real transactions are happening on crypto exchanges. 1040 00:50:54,140 --> 00:50:56,130 My thought about crypto exchanges-- 1041 00:50:56,130 --> 00:50:59,990 they'll not be 200 of them that much longer either. 1042 00:50:59,990 --> 00:51:02,060 We talked about Initial Coin Offerings. 1043 00:51:02,060 --> 00:51:05,670 Central takeaway is they're fundraising vehicles. 1044 00:51:05,670 --> 00:51:07,430 Whether you believe in or not believe 1045 00:51:07,430 --> 00:51:11,420 in the US and Canadian approach under the Howey Test, 1046 00:51:11,420 --> 00:51:16,760 but they help build a network, raise money in anticipation 1047 00:51:16,760 --> 00:51:19,985 the tokens are issued before it's functional. 1048 00:51:19,985 --> 00:51:24,110 I think the latest study was that-- 1049 00:51:24,110 --> 00:51:26,570 what was it-- 1 and 1/2 percent are actually functional, 1050 00:51:26,570 --> 00:51:29,210 and 98 and 1/2 percent are issued pre-functional. 1051 00:51:29,210 --> 00:51:31,550 So that's not a problem. 1052 00:51:31,550 --> 00:51:34,610 It's just-- that's the-- 1053 00:51:34,610 --> 00:51:36,830 means it's probably an investment vehicle, not 1054 00:51:36,830 --> 00:51:38,270 a utility token. 1055 00:51:38,270 --> 00:51:40,260 And that whole thing about utility token 1056 00:51:40,260 --> 00:51:46,010 versus security tokens, it stretched pretty darn thin. 1057 00:51:46,010 --> 00:51:48,680 We talked about public policy frameworks, 1058 00:51:48,680 --> 00:51:51,200 guarding against illicit activity, financial stability, 1059 00:51:51,200 --> 00:51:52,610 investing public. 1060 00:51:52,610 --> 00:51:56,090 My overall thought on this-- 1061 00:51:56,090 --> 00:51:59,210 and I accept that others will have other views-- 1062 00:51:59,210 --> 00:52:05,240 is that any new technology, if it grows to be economy-wide, 1063 00:52:05,240 --> 00:52:08,630 needs to live within public policy frameworks, 1064 00:52:08,630 --> 00:52:12,560 or society will change those public policy frameworks 1065 00:52:12,560 --> 00:52:14,190 modestly. 1066 00:52:14,190 --> 00:52:16,070 But you still want to protect the public 1067 00:52:16,070 --> 00:52:18,790 against core social goods. 1068 00:52:18,790 --> 00:52:20,590 It happened when the railroad came along. 1069 00:52:20,590 --> 00:52:23,230 It happened when telegram came along in the 19th century. 1070 00:52:23,230 --> 00:52:25,610 It happened with the internet. 1071 00:52:25,610 --> 00:52:27,770 Societies don't have technologies 1072 00:52:27,770 --> 00:52:33,870 fully outside of whatever social fabric we want to create. 1073 00:52:33,870 --> 00:52:35,850 And we do that social fabric stuff 1074 00:52:35,850 --> 00:52:40,770 through politics and parliaments and executives and regulators. 1075 00:52:40,770 --> 00:52:45,490 But we don't leave big parts of our economy outside of it. 1076 00:52:45,490 --> 00:52:47,545 This stuff is still small. 1077 00:52:47,545 --> 00:52:49,240 It's $100 billion. 1078 00:52:49,240 --> 00:52:54,240 The worldwide capital markets are over $300 trillion. 1079 00:52:54,240 --> 00:52:57,930 But it's gained a lot of public attention. 1080 00:52:57,930 --> 00:53:01,740 And particularly, we want to guard against illicit activity. 1081 00:53:01,740 --> 00:53:04,650 Most folks do. 1082 00:53:04,650 --> 00:53:05,790 Maybe not everybody. 1083 00:53:05,790 --> 00:53:07,310 But most do. 1084 00:53:07,310 --> 00:53:08,850 No, I recognize. 1085 00:53:08,850 --> 00:53:09,990 Hopefully, you all do. 1086 00:53:09,990 --> 00:53:10,800 Well, I don't know. 1087 00:53:13,320 --> 00:53:16,790 So we talked about the US Securities Law, the Howey Test. 1088 00:53:16,790 --> 00:53:21,090 It's not-- if you forget the Howey Test, it's totally fine. 1089 00:53:21,090 --> 00:53:24,570 But it's core to these initial coin offerings, 1090 00:53:24,570 --> 00:53:26,490 and it was that four-part test-- 1091 00:53:26,490 --> 00:53:28,500 investment of money in a common enterprise, 1092 00:53:28,500 --> 00:53:31,030 expectation of profits. 1093 00:53:31,030 --> 00:53:32,760 So if you get into some debate, and you 1094 00:53:32,760 --> 00:53:34,330 want to have some fun with somebody, 1095 00:53:34,330 --> 00:53:35,850 you can say you know the Howey Test. 1096 00:53:38,550 --> 00:53:40,390 But it's really more this. 1097 00:53:40,390 --> 00:53:42,450 I prefer the Duck Test. 1098 00:53:42,450 --> 00:53:44,280 Just use common sense. 1099 00:53:44,280 --> 00:53:47,460 This course has been more about critical reasoning skills 1100 00:53:47,460 --> 00:53:48,690 than anything. 1101 00:53:48,690 --> 00:53:51,450 So if somebody's trying to sell you something or hustle you 1102 00:53:51,450 --> 00:53:54,300 on something, and you are now a venture capitalist, 1103 00:53:54,300 --> 00:53:57,870 and you're investing, use the Duck Test. 1104 00:53:57,870 --> 00:53:59,430 You can even remember the Duck Test 1105 00:53:59,430 --> 00:54:01,770 about whether they're trying to sell you something, 1106 00:54:01,770 --> 00:54:04,090 and it has nothing to do with the law. 1107 00:54:04,090 --> 00:54:11,010 It just-- so where do I think crypto exchanges will go? 1108 00:54:11,010 --> 00:54:15,390 I think that they'll have to fix this whole custodial duty 1109 00:54:15,390 --> 00:54:16,110 stuff. 1110 00:54:16,110 --> 00:54:19,750 They might all use Fidelity's product. 1111 00:54:19,750 --> 00:54:24,780 Or they'll keep it in-house or spin it off. 1112 00:54:24,780 --> 00:54:27,790 I think they'll have to start complying with any money 1113 00:54:27,790 --> 00:54:28,290 laundering. 1114 00:54:28,290 --> 00:54:31,890 We saw one study that said a quarter of all exchanges 1115 00:54:31,890 --> 00:54:34,530 don't even have any AML. 1116 00:54:34,530 --> 00:54:38,970 I think two years from now, that will be way down. 1117 00:54:38,970 --> 00:54:42,150 I think margins will compress, and I 1118 00:54:42,150 --> 00:54:43,980 don't think there'll be 200 exchanges 1119 00:54:43,980 --> 00:54:45,840 two or three years from now. 1120 00:54:45,840 --> 00:54:48,570 There'll be five or 10 relevant exchanges 1121 00:54:48,570 --> 00:54:51,360 at some point in time. 1122 00:54:51,360 --> 00:54:52,942 On Initial Coin Offerings, I think 1123 00:54:52,942 --> 00:54:54,900 we're going to keep seeing a high failure rate. 1124 00:54:54,900 --> 00:54:57,150 There's still about 200 a month, but there'll probably 1125 00:54:57,150 --> 00:54:59,730 be fewer of them. 1126 00:54:59,730 --> 00:55:03,180 I think that there's going to be a lot of enforcement actions. 1127 00:55:03,180 --> 00:55:04,860 Occasionally, you'll see something-- 1128 00:55:04,860 --> 00:55:07,430 the Securities and Exchange Commission or some other-- 1129 00:55:07,430 --> 00:55:09,240 maybe it will be in Japan or elsewhere will 1130 00:55:09,240 --> 00:55:11,310 be bringing more cases. 1131 00:55:11,310 --> 00:55:14,670 And what will be interesting in 2019 is to watch, 1132 00:55:14,670 --> 00:55:17,580 does any of the big ones go live? 1133 00:55:17,580 --> 00:55:19,740 And I think this will be a real test. 1134 00:55:19,740 --> 00:55:21,990 Does Filecoin figure out what to do? 1135 00:55:21,990 --> 00:55:24,690 They raised a quarter of a billion dollars. 1136 00:55:24,690 --> 00:55:29,420 Telegram raised $1.7 billion. 1137 00:55:29,420 --> 00:55:31,430 Do they figure out what to do and make 1138 00:55:31,430 --> 00:55:34,040 a real live effort in 2019? 1139 00:55:34,040 --> 00:55:36,140 And it will test the economics. 1140 00:55:36,140 --> 00:55:38,060 And by the way, Eric, it might pull me more 1141 00:55:38,060 --> 00:55:40,640 towards where you are, right? 1142 00:55:40,640 --> 00:55:43,610 I mean Telegram's-- they got a lot of smart folks over there 1143 00:55:43,610 --> 00:55:45,320 working on this. 1144 00:55:45,320 --> 00:55:47,820 And by the way, Facebook has advertisements out 1145 00:55:47,820 --> 00:55:50,380 hiring people on blockchain technology. 1146 00:55:50,380 --> 00:55:53,280 So there's a lot of very big tech companies hiring 1147 00:55:53,280 --> 00:55:55,060 and recruiting in this space. 1148 00:55:55,060 --> 00:55:57,930 But I think 2019 will be interesting to see, what does-- 1149 00:55:57,930 --> 00:55:58,680 what is Telegram? 1150 00:55:58,680 --> 00:55:59,580 What is Filecoin? 1151 00:55:59,580 --> 00:56:03,000 What do some of the big, hopefully, well-meaning 1152 00:56:03,000 --> 00:56:06,990 companies do with these tokens? 1153 00:56:06,990 --> 00:56:10,650 Central banks-- mostly in the monitor phase. 1154 00:56:10,650 --> 00:56:12,948 But I'd say the thing to watch is Sweden. 1155 00:56:12,948 --> 00:56:14,490 It'll be really interesting what they 1156 00:56:14,490 --> 00:56:16,370 do with their E-krona project. 1157 00:56:16,370 --> 00:56:20,610 Do they go live and say, inspired by blockchain? 1158 00:56:20,610 --> 00:56:23,490 The E-krona project is not blockchain technology, 1159 00:56:23,490 --> 00:56:27,930 but it's completely inspired by Satoshi Nakamoto saying, well, 1160 00:56:27,930 --> 00:56:33,000 we, the central bank, Jihee's the central bank says, 1161 00:56:33,000 --> 00:56:36,240 we're going to do what Satoshi Nakamoto's-- 1162 00:56:36,240 --> 00:56:39,630 we're going to have a central bank digital token. 1163 00:56:39,630 --> 00:56:41,140 You can use it at Starbucks. 1164 00:56:41,140 --> 00:56:42,630 You can use it wherever. 1165 00:56:42,630 --> 00:56:43,410 I'd watch that. 1166 00:56:43,410 --> 00:56:47,690 I think Venezuela will probably fail on the other end. 1167 00:56:47,690 --> 00:56:48,780 But it'll be interesting. 1168 00:56:48,780 --> 00:56:54,720 There might be some distressed country that does something. 1169 00:56:54,720 --> 00:56:57,310 So how would I conclude? 1170 00:56:57,310 --> 00:56:58,680 I think blockchain does-- 1171 00:56:58,680 --> 00:57:01,110 this is my disclosure. 1172 00:57:01,110 --> 00:57:04,560 I think it does provide a real peer-to-peer alternative. 1173 00:57:04,560 --> 00:57:08,640 I think it's a legit live application. 1174 00:57:08,640 --> 00:57:11,860 It might not be the best, but is it better than some others? 1175 00:57:11,860 --> 00:57:14,550 Is it something you could use? 1176 00:57:14,550 --> 00:57:16,676 Verification and networking cost-- 1177 00:57:16,676 --> 00:57:19,110 I'm going to repeat on that-- 1178 00:57:19,110 --> 00:57:21,330 is the key. 1179 00:57:21,330 --> 00:57:23,520 But you always, I think, have to assess, is it 1180 00:57:23,520 --> 00:57:24,930 the right use case? 1181 00:57:24,930 --> 00:57:28,013 Is this use case the right use case? 1182 00:57:28,013 --> 00:57:29,430 But as we've talked, I think there 1183 00:57:29,430 --> 00:57:32,940 are real use cases more in permissioned 1184 00:57:32,940 --> 00:57:34,510 than permissionless. 1185 00:57:37,270 --> 00:57:38,110 Remember anything. 1186 00:57:38,110 --> 00:57:42,790 Money is just-- we all humans, the 7 billion 1187 00:57:42,790 --> 00:57:45,730 of us that live now, have to thank somebody 1188 00:57:45,730 --> 00:57:50,020 about 10,000 years ago, but came up with this technology called 1189 00:57:50,020 --> 00:57:51,370 money. 1190 00:57:51,370 --> 00:57:53,560 Store value, medium of exchange, unit 1191 00:57:53,560 --> 00:57:55,870 of account-- it's just a social construct, which 1192 00:57:55,870 --> 00:57:58,480 is a reason why I say, we could have a social construct 1193 00:57:58,480 --> 00:58:01,380 that is decentralized money. 1194 00:58:01,380 --> 00:58:04,330 I absolutely believe we could have that. 1195 00:58:04,330 --> 00:58:06,130 But we'd all have to accept it. 1196 00:58:06,130 --> 00:58:09,160 Where I have less optimism is that you'd 1197 00:58:09,160 --> 00:58:13,360 want to accept it for a closed, bounded ecosystem 1198 00:58:13,360 --> 00:58:15,340 only for file storage. 1199 00:58:15,340 --> 00:58:18,430 That's where I'm a little challenged as to, 1200 00:58:18,430 --> 00:58:23,930 what is the real economics of that? 1201 00:58:23,930 --> 00:58:25,820 Financial sector's characteristics 1202 00:58:25,820 --> 00:58:27,560 and challenges-- 1203 00:58:27,560 --> 00:58:29,080 is it good? 1204 00:58:29,080 --> 00:58:33,530 It's fertile ground, because finance is built on ledgers. 1205 00:58:33,530 --> 00:58:35,090 And finance is big-- 1206 00:58:35,090 --> 00:58:38,440 7 and 1/2 percent of our economy. 1207 00:58:38,440 --> 00:58:41,210 And around the globe, it's anywhere from 5% 1208 00:58:41,210 --> 00:58:43,420 to probably 8% of those economies. 1209 00:58:43,420 --> 00:58:45,510 Most other countries, it's less than the US, 1210 00:58:45,510 --> 00:58:49,190 but it's still-- it's big. 1211 00:58:49,190 --> 00:58:50,690 Incumbents are, of course, looking 1212 00:58:50,690 --> 00:58:52,970 at permissioned blockchains. 1213 00:58:52,970 --> 00:58:54,740 And unfortunately, crypto finance 1214 00:58:54,740 --> 00:58:57,800 has a bunch of scams and frauds. 1215 00:58:57,800 --> 00:59:01,860 And that hurts the development. 1216 00:59:01,860 --> 00:59:05,730 But I think adoption will come, but it 1217 00:59:05,730 --> 00:59:08,070 rests on solving a bunch of technical issues, 1218 00:59:08,070 --> 00:59:11,800 like the Digital Currency Initiative's trying to solve. 1219 00:59:11,800 --> 00:59:14,560 What are the commercial use cases and the public policy? 1220 00:59:14,560 --> 00:59:16,110 I wouldn't lead with public policy. 1221 00:59:16,110 --> 00:59:18,380 I don't think that's the main issue here. 1222 00:59:18,380 --> 00:59:21,200 But if this is ever going to be anything, 1223 00:59:21,200 --> 00:59:24,400 it's got to get over that. 1224 00:59:24,400 --> 00:59:26,920 So mostly, it's probably going to be a catalyst for change, 1225 00:59:26,920 --> 00:59:31,710 more than actual real thoughts. 1226 00:59:31,710 --> 00:59:32,722 Questions? 1227 00:59:35,820 --> 00:59:37,620 Thoughts? 1228 00:59:37,620 --> 00:59:38,185 Eric. 1229 00:59:38,185 --> 00:59:41,653 AUDIENCE: On the third point, I-- 1230 00:59:41,653 --> 00:59:43,320 GARY GENSLER: Use cases must address why 1231 00:59:43,320 --> 00:59:45,060 versus a traditional database. 1232 00:59:45,060 --> 00:59:46,530 AUDIENCE: I always had the feeling 1233 00:59:46,530 --> 00:59:49,470 that it seemed the permissioned space, where 1234 00:59:49,470 --> 00:59:54,190 you have the most or the biggest urge to prove where-- 1235 00:59:54,190 --> 01:00:00,630 if your submission needs a blockchain solution as opposed 1236 01:00:00,630 --> 01:00:01,925 to a traditional-- 1237 01:00:01,925 --> 01:00:05,040 I mean making a case against the centralized database 1238 01:00:05,040 --> 01:00:08,755 is kind of easy, versus a distributed database. 1239 01:00:08,755 --> 01:00:11,210 I think it's a little bit tough. 1240 01:00:11,210 --> 01:00:13,987 But in the permissionless space, I 1241 01:00:13,987 --> 01:00:15,570 think you don't have to struggle much, 1242 01:00:15,570 --> 01:00:18,810 because it's absolutely justified that you can-- 1243 01:00:18,810 --> 01:00:23,010 because you're dealing with a decentralized configuration, 1244 01:00:23,010 --> 01:00:27,900 where there's no one in charge of guaranteeing 1245 01:00:27,900 --> 01:00:30,020 that we have a-- 1246 01:00:30,020 --> 01:00:32,310 GARY GENSLER: So Eric's raising the question, 1247 01:00:32,310 --> 01:00:34,480 why not use a traditional database? 1248 01:00:34,480 --> 01:00:36,120 Why ever use a permissioned-- 1249 01:00:36,120 --> 01:00:37,620 AUDIENCE: In the permissioned space, 1250 01:00:37,620 --> 01:00:40,080 because permissionless is perfectly justified. 1251 01:00:40,080 --> 01:00:41,550 But in the permissioned space, I'm 1252 01:00:41,550 --> 01:00:44,500 not saying it's not justified. 1253 01:00:44,500 --> 01:00:48,840 The point is being that in our project, 1254 01:00:48,840 --> 01:00:51,810 we argued for a permissioned system. 1255 01:00:51,810 --> 01:00:53,910 We made some justifications in that space, 1256 01:00:53,910 --> 01:00:59,970 so the point is that I just wanted to get your thoughts so 1257 01:00:59,970 --> 01:01:01,760 that in the permissionless spaces, 1258 01:01:01,760 --> 01:01:04,140 you don't have to struggle much in that. 1259 01:01:04,140 --> 01:01:08,800 GARY GENSLER: Permissionless blockchain technology, 1260 01:01:08,800 --> 01:01:10,480 looking forward a number of years, 1261 01:01:10,480 --> 01:01:13,570 looking forward to when some of the scalability and performance 1262 01:01:13,570 --> 01:01:17,540 gets better, can give you verification costs. 1263 01:01:17,540 --> 01:01:22,720 And it's censorship-resistant or practically 1264 01:01:22,720 --> 01:01:27,010 censorship-resistant, unless somebody does a 51% attack. 1265 01:01:27,010 --> 01:01:31,120 But so I think that the truly decentralized distributed 1266 01:01:31,120 --> 01:01:36,480 permissionless system has a lot of-- a lot to offer. 1267 01:01:36,480 --> 01:01:39,710 But it comes with a lot of costs, too, and challenges-- 1268 01:01:39,710 --> 01:01:44,240 hard to do the governance and hard to get coordinated action. 1269 01:01:44,240 --> 01:01:46,700 And as you've heard me say, I think 1270 01:01:46,700 --> 01:01:49,400 there's only probably going to be a very limited use 1271 01:01:49,400 --> 01:01:56,400 cases for native tokens, native product-specific tokens 1272 01:01:56,400 --> 01:02:03,690 like a Filecoin or a restaurant review coin or a health care 1273 01:02:03,690 --> 01:02:04,740 coin. 1274 01:02:04,740 --> 01:02:08,070 I have-- I have my doubts about that. 1275 01:02:08,070 --> 01:02:11,820 An economy-wide coin, a cryptocurrency like Bitcoin-- 1276 01:02:11,820 --> 01:02:14,190 I think there's more use, particularly 1277 01:02:14,190 --> 01:02:17,520 if you have a weak central bank, a weak fiscal policy. 1278 01:02:17,520 --> 01:02:22,620 A country like Venezuela in the future might adopt something. 1279 01:02:22,620 --> 01:02:24,840 I think there's a lot to be said for permissionless, 1280 01:02:24,840 --> 01:02:28,590 but permissioned systems versus traditional databases, I think, 1281 01:02:28,590 --> 01:02:30,180 also gives something. 1282 01:02:30,180 --> 01:02:33,150 And I'll use the Australian stock exchange for one, 1283 01:02:33,150 --> 01:02:34,950 but they know their customers. 1284 01:02:34,950 --> 01:02:35,910 They know they have-- 1285 01:02:35,910 --> 01:02:38,640 I think the number was 77 members. 1286 01:02:38,640 --> 01:02:42,510 They can take their database and decentralize it and put it 1287 01:02:42,510 --> 01:02:47,730 on all 77 member companies, if that's the right number, 1288 01:02:47,730 --> 01:02:51,180 and say, you're now sharing that. 1289 01:02:51,180 --> 01:02:53,610 And so then there's no single point of failure, 1290 01:02:53,610 --> 01:02:56,370 and we can lower what's called reconciliation costs, 1291 01:02:56,370 --> 01:02:59,370 because you all needed some of that data anyway. 1292 01:02:59,370 --> 01:03:04,620 The historic data-keeping, the historic nature of it 1293 01:03:04,620 --> 01:03:07,230 came up with each of those 77 members 1294 01:03:07,230 --> 01:03:09,370 having individual databases. 1295 01:03:09,370 --> 01:03:12,540 So I could see that there is a real benefit in probably 1296 01:03:12,540 --> 01:03:15,810 efficiency and lowering costs, which 1297 01:03:15,810 --> 01:03:18,650 means lowering costs of verification, 1298 01:03:18,650 --> 01:03:20,400 lowering networking costs, because there's 1299 01:03:20,400 --> 01:03:23,250 77 members, trade finance. 1300 01:03:23,250 --> 01:03:25,030 I could see some of that as well. 1301 01:03:25,030 --> 01:03:27,240 So that-- but you might be right. 1302 01:03:27,240 --> 01:03:28,920 Oracle might be able to just adopt 1303 01:03:28,920 --> 01:03:30,290 that in a traditional database. 1304 01:03:30,290 --> 01:03:30,865 Sabrina. 1305 01:03:30,865 --> 01:03:32,490 AUDIENCE: Yeah, so it seems like you're 1306 01:03:32,490 --> 01:03:35,800 describing that they have some form of distributed consensus. 1307 01:03:35,800 --> 01:03:37,390 And that's the main reason, right? 1308 01:03:37,390 --> 01:03:38,836 But there's-- 1309 01:03:38,836 --> 01:03:40,890 GARY GENSLER: And distributed the data 1310 01:03:40,890 --> 01:03:43,980 stored on multiple nodes, because it's multiple copies 1311 01:03:43,980 --> 01:03:45,040 of the same data. 1312 01:03:45,040 --> 01:03:47,580 AUDIENCE: Right, but yeah. 1313 01:03:47,580 --> 01:03:50,690 There's-- I mean if that still can be implemented without 1314 01:03:50,690 --> 01:03:51,770 watching. 1315 01:03:51,770 --> 01:03:54,456 I'm playing the Aleem today, since he's not here. 1316 01:03:54,456 --> 01:03:55,517 But there's no-- 1317 01:03:55,517 --> 01:03:56,600 GARY GENSLER: No, but we-- 1318 01:03:56,600 --> 01:03:59,670 AUDIENCE: That's been a long solved problem of how consensus 1319 01:03:59,670 --> 01:04:01,360 algorithms that-- or you can replicate 1320 01:04:01,360 --> 01:04:06,380 state on the same state on both the machines 1321 01:04:06,380 --> 01:04:09,490 and tolerate some [INAUDIBLE] number of failures. 1322 01:04:09,490 --> 01:04:14,470 So I personally don't really see the use of blockchain 1323 01:04:14,470 --> 01:04:18,990 in permissioned systems, but I do see a lot of potential 1324 01:04:18,990 --> 01:04:21,780 for permissionless systems, just because those algorithms 1325 01:04:21,780 --> 01:04:24,540 already exist out there to do [INAUDIBLE].. 1326 01:04:24,540 --> 01:04:26,550 GARY GENSLER: So what Sabrina's raising-- 1327 01:04:26,550 --> 01:04:28,410 you can take the Sabrina seat. 1328 01:04:28,410 --> 01:04:34,060 You're-- Sabrina's a master's student at computer science 1329 01:04:34,060 --> 01:04:38,160 at MIT, so she probably knows what she's talking about-- 1330 01:04:38,160 --> 01:04:41,910 is really saying, well, you can do a lot of that. 1331 01:04:41,910 --> 01:04:44,640 You can have distributed databases 1332 01:04:44,640 --> 01:04:47,250 without blockchain technology. 1333 01:04:47,250 --> 01:04:49,380 You can have-- and there's ways that 1334 01:04:49,380 --> 01:04:52,380 this Byzantine-fought tolerance has been addressed 1335 01:04:52,380 --> 01:04:55,860 in distributed databases that you don't 1336 01:04:55,860 --> 01:04:57,510 need blockchain technology. 1337 01:04:57,510 --> 01:05:02,055 The concept of a block, commitment hash, block, 1338 01:05:02,055 --> 01:05:06,100 a commitment hash, block technology. 1339 01:05:06,100 --> 01:05:07,240 And I accept that. 1340 01:05:07,240 --> 01:05:10,890 But I'm not going to accept it to say 1341 01:05:10,890 --> 01:05:13,830 that there's no space for permissioned blockchain 1342 01:05:13,830 --> 01:05:15,510 technology. 1343 01:05:15,510 --> 01:05:17,880 What there is space for is the 80 of you 1344 01:05:17,880 --> 01:05:21,760 in this room to have different opinions on all of this. 1345 01:05:21,760 --> 01:05:22,765 You want to close out? 1346 01:05:22,765 --> 01:05:24,640 I want to-- I have one more slide I want to-- 1347 01:05:24,640 --> 01:05:26,628 AUDIENCE: [INAUDIBLE] 1348 01:05:26,628 --> 01:05:28,670 GARY GENSLER: So this is, of course, about money, 1349 01:05:28,670 --> 01:05:31,760 so I went back and found a quote from Benjamin Franklin 1350 01:05:31,760 --> 01:05:32,750 that I like. 1351 01:05:32,750 --> 01:05:36,320 And to me, it's about a concept that I hope you all live by. 1352 01:05:36,320 --> 01:05:38,920 It's paying it forward. 1353 01:05:38,920 --> 01:05:43,520 Paying forward is partly because all of us in this room 1354 01:05:43,520 --> 01:05:44,360 are privileged. 1355 01:05:44,360 --> 01:05:51,170 We're at MIT or Harvard or Wellesley, but we're at MIT. 1356 01:05:51,170 --> 01:05:52,970 And so what did Ben Franklin say? 1357 01:05:52,970 --> 01:05:55,430 And it does close on money. 1358 01:05:55,430 --> 01:05:57,260 "I do not pretend to give such a deed. 1359 01:05:57,260 --> 01:05:59,750 I only lend it to you. 1360 01:05:59,750 --> 01:06:02,540 When you meet another honest man in similar distress, 1361 01:06:02,540 --> 01:06:06,980 you must pay me by lending this sum to him 1362 01:06:06,980 --> 01:06:11,390 and joining him to discharge the debt by a like operation," 1363 01:06:11,390 --> 01:06:14,770 paying it forward, basically. 1364 01:06:14,770 --> 01:06:17,020 "He should be able and shall meet another opportunity, 1365 01:06:17,020 --> 01:06:20,690 I hope, that this may, though many hands." 1366 01:06:20,690 --> 01:06:22,910 So Benjamin Franklin, over 200 years ago, 1367 01:06:22,910 --> 01:06:24,770 he closes, this is a trick of mine 1368 01:06:24,770 --> 01:06:28,540 of doing a deal good with a little money.