1 00:00:00,266 --> 00:00:00,973 The age of AI 2 00:00:01,346 --> 00:00:03,412 Your fund manager may be redundant 3 00:00:04,078 --> 00:00:04,386 Hello 4 00:00:04,679 --> 00:00:05,106 Hello everyone 5 00:00:05,279 --> 00:00:06,205 I am Wang Lijie 6 00:00:06,479 --> 00:00:09,645 Today we tackle a question keeping Wall Street awake 7 00:00:10,310 --> 00:00:11,017 In 2024 8 00:00:11,310 --> 00:00:13,398 Citibank released a report 9 00:00:13,470 --> 00:00:17,877 Saying 54 percent of banking jobs have high automation potential 10 00:00:18,031 --> 00:00:18,697 Fifty four percent 11 00:00:19,031 --> 00:00:20,077 More than half 12 00:00:20,310 --> 00:00:22,557 But look at the Wall Street layoff data 13 00:00:22,591 --> 00:00:24,589 People actually displaced by AI 14 00:00:24,591 --> 00:00:26,118 Are still almost zero 15 00:00:26,350 --> 00:00:28,118 So what explains this huge gap 16 00:00:28,231 --> 00:00:30,510 Is Citi manufacturing anxiety 17 00:00:30,510 --> 00:00:32,178 Or is this the calm before the storm 18 00:00:33,510 --> 00:00:36,998 As an angel investor in over 300 companies 19 00:00:37,190 --> 00:00:39,757 This is not merely somebody else problem 20 00:00:39,950 --> 00:00:41,998 I use AI every day 21 00:00:42,031 --> 00:00:44,709 To analyze data and screen deals 22 00:00:44,711 --> 00:00:46,077 And track market signals 23 00:00:46,349 --> 00:00:47,150 Honestly 24 00:00:47,151 --> 00:00:50,077 It already does some things faster and better than I do 25 00:00:50,151 --> 00:00:52,377 So what value do human investors still have 26 00:00:52,870 --> 00:00:55,218 Could I also be replaced by AI one day 27 00:00:55,711 --> 00:00:57,762 Today we will break this question down completely 28 00:00:57,994 --> 00:00:58,701 Let me give you the conclusion first 29 00:00:59,075 --> 00:01:00,921 Then we will unpack the logic 30 00:01:01,034 --> 00:01:01,621 Here is the conclusion 31 00:01:02,155 --> 00:01:04,201 AI will not replace fund managers 32 00:01:04,234 --> 00:01:06,321 But it will force the species of fund managers 33 00:01:06,354 --> 00:01:10,041 Into a sharp evolutionary divide 34 00:01:10,195 --> 00:01:12,562 Some will become extraordinarily valuable 35 00:01:12,634 --> 00:01:15,082 Others will become worthless 36 00:01:15,235 --> 00:01:17,273 What determines which side you land on 37 00:01:17,274 --> 00:01:19,002 Is not whether you can use AI 38 00:01:19,074 --> 00:01:22,801 It is whether you have something AI can never learn 39 00:01:23,395 --> 00:01:23,582 All right 40 00:01:23,955 --> 00:01:26,202 Now let us unpack the logic layer by layer 41 00:01:26,395 --> 00:01:27,381 Starting with the data 42 00:01:27,634 --> 00:01:28,842 Then the underlying causes 43 00:01:28,875 --> 00:01:31,282 And finally how you should think about it 44 00:01:32,331 --> 00:01:32,757 Layer one 45 00:01:33,210 --> 00:01:34,858 What happened in the data 46 00:01:35,051 --> 00:01:36,138 In 2025 47 00:01:36,210 --> 00:01:39,257 An empirical study in the Journal of Future Business 48 00:01:39,331 --> 00:01:41,370 Compared AI driven funds 49 00:01:41,370 --> 00:01:43,049 With human managed funds 50 00:01:43,051 --> 00:01:45,097 Across different market conditions 51 00:01:45,251 --> 00:01:49,658 Using global equity funds focused on ESG and technology 52 00:01:49,971 --> 00:01:53,977 Researchers studied the 2022 bear market and 2024 bull market 53 00:01:54,090 --> 00:01:58,578 Using risk adjusted measures like Sharpe ratio and Jensen alpha 54 00:01:58,811 --> 00:01:59,757 The results were fascinating 55 00:02:00,251 --> 00:02:02,097 In the 2022 bear market 56 00:02:02,251 --> 00:02:05,578 AI funds lost over 40 percent less than human funds 57 00:02:05,730 --> 00:02:08,778 Jensen alpha was positive for AI and negative for humans 58 00:02:09,490 --> 00:02:10,278 AI won 59 00:02:10,490 --> 00:02:11,977 And it won decisively 60 00:02:12,171 --> 00:02:14,417 But in the 2024 bull market 61 00:02:14,530 --> 00:02:16,058 The picture completely reversed 62 00:02:16,250 --> 00:02:19,178 Human funds had a Jensen alpha of positive 5 44 63 00:02:19,331 --> 00:02:21,778 While AI funds had negative 7 93 64 00:02:22,051 --> 00:02:23,357 Humans did not merely edge ahead 65 00:02:23,771 --> 00:02:25,377 They won by a landslide 66 00:02:25,822 --> 00:02:26,329 Why 67 00:02:26,863 --> 00:02:28,489 What is the core task in a bear market 68 00:02:28,982 --> 00:02:29,730 Cutting losses 69 00:02:29,982 --> 00:02:30,869 Managing risk 70 00:02:30,942 --> 00:02:33,149 And avoiding foolish decisions in a panic 71 00:02:33,382 --> 00:02:35,389 That is exactly where algorithms excel 72 00:02:35,942 --> 00:02:37,670 Algorithms have no adrenaline 73 00:02:37,702 --> 00:02:39,670 Their hands do not shake in a crash 74 00:02:39,822 --> 00:02:42,750 They do not hold forever because selling hurts 75 00:02:43,262 --> 00:02:46,589 The greatest enemy of human investors is never the market 76 00:02:46,702 --> 00:02:48,470 It is their own amygdala 77 00:02:48,822 --> 00:02:50,829 AI simply does not have that flaw 78 00:02:51,327 --> 00:02:53,033 But what is the core task in a bull market 79 00:02:53,486 --> 00:02:55,133 Judging turning points in a trend 80 00:02:55,246 --> 00:02:58,494 Making qualitative judgments amid extreme ambiguity 81 00:02:58,646 --> 00:03:02,133 Knowing from a founders eyes whether the venture can work 82 00:03:02,406 --> 00:03:05,006 These tasks do not need more data 83 00:03:05,007 --> 00:03:07,013 They need intuition from less data 84 00:03:07,406 --> 00:03:09,093 I have made angel investments for over a decade 85 00:03:09,127 --> 00:03:10,614 My most important decisions 86 00:03:10,646 --> 00:03:14,214 Investing in EHang Hesai Dobot and Xunce 87 00:03:14,327 --> 00:03:16,793 Which one came from a spreadsheet model 88 00:03:17,150 --> 00:03:17,949 None 89 00:03:18,127 --> 00:03:20,406 Those companies had no revenue at the time 90 00:03:20,406 --> 00:03:21,234 What could you calculate 91 00:03:21,686 --> 00:03:23,533 The only thing you could calculate was probability 92 00:03:23,686 --> 00:03:25,853 But probability itself is subjective 93 00:03:25,926 --> 00:03:28,494 It is a bet on how the world works 94 00:03:29,287 --> 00:03:29,954 That kind of bet 95 00:03:30,127 --> 00:03:31,942 Is beyond AI for now 96 00:03:32,295 --> 00:03:32,761 Layer two 97 00:03:33,134 --> 00:03:34,842 Why does AI struggle in bull markets 98 00:03:35,374 --> 00:03:38,062 The answer lies somewhere deeper 99 00:03:38,374 --> 00:03:39,541 Have you ever wondered 100 00:03:39,654 --> 00:03:41,442 How a bull market actually forms 101 00:03:42,055 --> 00:03:43,942 It is not simply improving fundamentals 102 00:03:44,215 --> 00:03:46,253 Often the fundamentals have barely moved 103 00:03:46,254 --> 00:03:47,622 Yet the stock price has already soared 104 00:03:47,934 --> 00:03:51,581 A bull market begins when people believe the same story 105 00:03:51,895 --> 00:03:54,622 The story might be that AI will change everything 106 00:03:54,775 --> 00:03:57,222 Or that the Chinese economy will take off again 107 00:03:57,414 --> 00:03:59,102 Or that interest rates have peaked 108 00:03:59,335 --> 00:04:01,454 What matters is not whether the story is true 109 00:04:01,455 --> 00:04:04,502 But whether more people are starting to believe it 110 00:04:05,602 --> 00:04:07,609 Soros called this reflexivity 111 00:04:08,202 --> 00:04:08,869 What does that mean 112 00:04:09,561 --> 00:04:12,848 Market participants change the market through their beliefs 113 00:04:13,322 --> 00:04:14,869 Everyone believes AI stocks will rise 114 00:04:15,202 --> 00:04:16,568 So they buy AI stocks 115 00:04:16,841 --> 00:04:18,528 And AI stocks really do rise 116 00:04:18,722 --> 00:04:20,688 The rise validates the belief 117 00:04:20,761 --> 00:04:21,828 More people pile in 118 00:04:22,042 --> 00:04:22,849 And prices keep rising 119 00:04:23,282 --> 00:04:24,588 The story changes behavior 120 00:04:25,001 --> 00:04:26,609 Behavior changes reality 121 00:04:26,801 --> 00:04:28,809 And reality reinforces the story 122 00:04:29,162 --> 00:04:31,049 It is a positive feedback loop 123 00:04:31,681 --> 00:04:32,588 Inside this loop 124 00:04:32,962 --> 00:04:34,049 Nothing 125 00:04:34,201 --> 00:04:36,489 Can be predicted from historical data 126 00:04:36,681 --> 00:04:40,625 Because it starts with a new story appearing from nowhere 127 00:04:40,979 --> 00:04:42,746 AI can recognize patterns 128 00:04:42,938 --> 00:04:46,585 But those are statistical patterns in historical data 129 00:04:46,859 --> 00:04:49,546 The biggest opportunities in a bull market 130 00:04:49,578 --> 00:04:52,625 Often come from narratives history has never seen 131 00:04:52,979 --> 00:04:54,666 Let me give you a concrete example 132 00:04:54,938 --> 00:04:56,186 In early 2023 133 00:04:56,258 --> 00:04:58,625 Nvidia shares were still a little over 100 dollars 134 00:04:58,859 --> 00:05:01,585 Look at any quantitative model from that time 135 00:05:01,618 --> 00:05:04,226 Or any AI system based on historical data 136 00:05:04,378 --> 00:05:08,505 None would predict the stock tripling within a year 137 00:05:08,899 --> 00:05:09,405 Why 138 00:05:09,899 --> 00:05:14,065 Because GPUs had never shown that growth pattern in 20 years 139 00:05:14,339 --> 00:05:16,664 There was no precedent in the data 140 00:05:17,097 --> 00:05:18,885 But what did human investors see 141 00:05:19,457 --> 00:05:23,265 They saw a new computing paradigm taking shape 142 00:05:23,657 --> 00:05:24,524 They saw 143 00:05:24,738 --> 00:05:26,464 The rise of large language models 144 00:05:26,537 --> 00:05:29,625 Would drive exponential growth in computing demand 145 00:05:29,818 --> 00:05:32,784 That judgment did not come from a stock chart 146 00:05:32,818 --> 00:05:36,104 It came from intuition about technology trends 147 00:05:36,418 --> 00:05:38,464 The AI rally of 2023 148 00:05:38,577 --> 00:05:42,745 Had no signal in any dataset from 2022 149 00:05:42,938 --> 00:05:45,745 Because it did not grow out of data 150 00:05:45,777 --> 00:05:49,024 It grew out of the collective human imagination 151 00:05:49,378 --> 00:05:52,945 The first thing human investors can do that AI cannot 152 00:05:52,978 --> 00:05:54,664 Is build a new narrative 153 00:05:55,178 --> 00:05:57,584 Not find patterns in old data 154 00:05:57,657 --> 00:06:00,065 But see a future that does not yet exist 155 00:06:00,178 --> 00:06:01,534 And then bet on it 156 00:06:02,126 --> 00:06:02,713 Layer three 157 00:06:03,207 --> 00:06:04,873 What is Wall Street really afraid of 158 00:06:05,326 --> 00:06:07,453 They are not afraid of being replaced by AI 159 00:06:07,806 --> 00:06:09,773 They are afraid of being compressed by AI 160 00:06:10,010 --> 00:06:10,474 What does that mean 161 00:06:11,087 --> 00:06:11,965 Consider an example 162 00:06:12,518 --> 00:06:14,398 In the past an investment banking analyst 163 00:06:14,398 --> 00:06:19,205 Spent 12 hours a day extracting data modeling and writing reports 164 00:06:19,359 --> 00:06:21,845 One person could cover at most 15 companies 165 00:06:22,119 --> 00:06:26,405 Now AI completes three days of data work in five minutes 166 00:06:26,559 --> 00:06:29,085 The same analyst can cover 60 companies 167 00:06:29,398 --> 00:06:31,165 Productivity has quadrupled 168 00:06:31,438 --> 00:06:32,665 What do you think happens next 169 00:06:33,159 --> 00:06:34,958 You do not need four times as many analysts 170 00:06:34,958 --> 00:06:36,725 You need only one quarter as many 171 00:06:37,198 --> 00:06:39,665 Executives at major Wall Street banks openly admit 172 00:06:40,119 --> 00:06:42,398 AI will not eliminate the analyst role 173 00:06:42,398 --> 00:06:45,366 But it will compress the work of five people into one job 174 00:06:45,638 --> 00:06:46,986 So where do the other four go 175 00:06:47,653 --> 00:06:50,660 That is the real meaning of the Citi figure of 54 percent 176 00:06:51,013 --> 00:06:52,780 Your job does not disappear 177 00:06:52,814 --> 00:06:54,980 Someone else absorbs it 178 00:06:55,134 --> 00:06:56,740 The one person left standing 179 00:06:56,814 --> 00:06:58,573 Will not be the best at using AI 180 00:06:58,573 --> 00:07:01,021 It will be the person best at what AI cannot do 181 00:07:01,174 --> 00:07:01,761 Layer four 182 00:07:02,134 --> 00:07:03,840 What can AI never do 183 00:07:04,294 --> 00:07:05,657 I can think of three things 184 00:07:05,888 --> 00:07:06,569 First 185 00:07:06,569 --> 00:07:09,537 Making decisions with severely incomplete information 186 00:07:09,889 --> 00:07:11,169 When I invest in early stage ventures 187 00:07:11,170 --> 00:07:13,336 I often have only a one page business plan 188 00:07:13,369 --> 00:07:15,256 Sometimes there is not even a product 189 00:07:15,609 --> 00:07:18,009 Faced with such scarce information 190 00:07:18,009 --> 00:07:19,729 AI can only say there is not enough data 191 00:07:19,730 --> 00:07:20,737 No judgment is possible 192 00:07:21,129 --> 00:07:24,776 But real investment opportunities hide where data is scarce 193 00:07:25,090 --> 00:07:26,117 Once the data is complete 194 00:07:26,330 --> 00:07:28,586 The excess returns are gone 195 00:07:28,818 --> 00:07:29,125 Second 196 00:07:29,339 --> 00:07:30,825 Bearing real consequences 197 00:07:31,219 --> 00:07:32,577 AI gives you advice 198 00:07:32,579 --> 00:07:34,466 But loses nothing when it is wrong 199 00:07:34,699 --> 00:07:37,546 A human fund manager may put everything on the line 200 00:07:37,579 --> 00:07:39,346 For a contrarian judgment 201 00:07:39,418 --> 00:07:41,385 That proves correct three years later 202 00:07:41,699 --> 00:07:44,786 Along the way they endure pressure loneliness and torment 203 00:07:44,858 --> 00:07:47,026 That suffering itself creates excess returns 204 00:07:47,459 --> 00:07:48,945 Imagine what that feels like 205 00:07:49,219 --> 00:07:50,697 Everyone says you are wrong 206 00:07:50,699 --> 00:07:52,378 Your partners question you 207 00:07:52,378 --> 00:07:54,026 Your investors pressure you 208 00:07:54,219 --> 00:07:55,897 And you stand there alone 209 00:07:55,899 --> 00:07:56,885 Relying not on data 210 00:07:57,178 --> 00:07:58,106 But conviction 211 00:07:58,298 --> 00:08:00,625 That pain itself is the barrier 212 00:08:00,858 --> 00:08:02,385 Without that barrier 213 00:08:02,418 --> 00:08:04,257 Everyone would do the same thing 214 00:08:04,258 --> 00:08:06,542 And excess returns would disappear 215 00:08:06,934 --> 00:08:07,282 Third 216 00:08:07,734 --> 00:08:09,261 Building trust with people 217 00:08:09,775 --> 00:08:12,181 Investing has never been purely a math problem 218 00:08:12,494 --> 00:08:15,581 Especially in private equity venture capital and angel investing 219 00:08:15,814 --> 00:08:17,022 You are investing in people 220 00:08:17,575 --> 00:08:19,701 My best investments over the years 221 00:08:19,775 --> 00:08:23,542 Were all built on deep interpersonal trust 222 00:08:24,015 --> 00:08:26,342 When a founders company was close to collapse 223 00:08:26,494 --> 00:08:28,381 They called me at two in the morning 224 00:08:28,575 --> 00:08:30,862 Not because my fund had the highest return 225 00:08:31,095 --> 00:08:32,862 But because they knew I would answer 226 00:08:32,895 --> 00:08:33,441 I would listen 227 00:08:33,895 --> 00:08:35,462 And help them find a way forward 228 00:08:35,854 --> 00:08:38,982 That relationship is earned through time and sincerity 229 00:08:39,334 --> 00:08:40,661 No matter how smart AI becomes 230 00:08:40,694 --> 00:08:43,862 It has no personhood you can call at two in the morning 231 00:08:44,375 --> 00:08:47,901 Trust is something only humans can transmit to one another 232 00:08:48,775 --> 00:08:50,381 You can trust a tool 233 00:08:50,495 --> 00:08:53,788 But you do not confide in a tool when you are desperate 234 00:08:54,260 --> 00:08:54,727 Layer five 235 00:08:55,060 --> 00:08:55,888 What about retail investors 236 00:08:56,500 --> 00:08:57,727 We have discussed professional investors 237 00:08:57,981 --> 00:08:59,068 Now let us discuss retail investors 238 00:08:59,300 --> 00:09:00,827 For retail investors 239 00:09:00,861 --> 00:09:03,148 The impact of AI is actually good news 240 00:09:03,323 --> 00:09:03,808 Why 241 00:09:04,260 --> 00:09:06,500 The biggest reason retail investors lose money 242 00:09:06,500 --> 00:09:08,140 Is never a lack of information 243 00:09:08,140 --> 00:09:09,989 It is too much emotion 244 00:09:10,582 --> 00:09:13,350 Think about whether this has happened to you 245 00:09:13,542 --> 00:09:14,769 The market rises for three straight days 246 00:09:15,062 --> 00:09:15,769 You get restless 247 00:09:15,902 --> 00:09:16,749 And chase the rally 248 00:09:16,983 --> 00:09:18,129 Then it falls on the fourth day 249 00:09:18,302 --> 00:09:18,889 You panic 250 00:09:19,103 --> 00:09:19,790 And sell 251 00:09:20,103 --> 00:09:22,029 Then it rises again on the fifth day 252 00:09:22,302 --> 00:09:23,382 After all that trading 253 00:09:23,383 --> 00:09:25,590 You earn ten percent less than by doing nothing 254 00:09:26,103 --> 00:09:26,810 It is not because you are foolish 255 00:09:27,343 --> 00:09:28,422 It is because your brain 256 00:09:28,422 --> 00:09:31,341 Evolved survival instincts in the primal jungle 257 00:09:31,343 --> 00:09:33,029 That fire wildly in the stock market 258 00:09:33,422 --> 00:09:35,302 You feel anxious when others profit 259 00:09:35,302 --> 00:09:37,149 And afraid when you lose money 260 00:09:37,383 --> 00:09:40,310 Those emotions could save your life in ancient times 261 00:09:40,343 --> 00:09:42,752 But only lose you money in the stock market 262 00:09:43,065 --> 00:09:46,344 Chasing rallies panic selling overtrading and confirmation bias 263 00:09:46,346 --> 00:09:49,913 Cost retail investors one or two percentage points each year 264 00:09:50,106 --> 00:09:51,545 Compounded over 30 years 265 00:09:51,545 --> 00:09:53,793 That erases 30 percent of your wealth 266 00:09:54,305 --> 00:09:56,313 If AI can help you control your impulses 267 00:09:56,425 --> 00:09:58,545 Stopping you when you want to sell in panic 268 00:09:58,545 --> 00:10:00,713 And warning you when you want to chase a high 269 00:10:00,746 --> 00:10:05,112 It may create more value than any star fund manager 270 00:10:05,305 --> 00:10:07,184 Vanguard conducted a study 271 00:10:07,185 --> 00:10:08,632 Called Advisor Alpha 272 00:10:08,785 --> 00:10:09,665 It found 273 00:10:09,665 --> 00:10:13,392 A financial advisor can add up to about three percent 274 00:10:14,266 --> 00:10:16,433 The largest part of that three percent 275 00:10:16,545 --> 00:10:17,745 One point five percent 276 00:10:17,746 --> 00:10:19,184 Does not come from stock picking 277 00:10:19,185 --> 00:10:20,573 Or market timing 278 00:10:20,986 --> 00:10:23,232 It comes from behavioral coaching 279 00:10:23,545 --> 00:10:26,842 Pulling you back just before you do something foolish 280 00:10:27,114 --> 00:10:29,154 AI can now do that 281 00:10:29,155 --> 00:10:30,502 It is available 24 hours a day 282 00:10:30,675 --> 00:10:31,633 It never gets tired 283 00:10:31,635 --> 00:10:32,761 It never gets impatient 284 00:10:32,795 --> 00:10:35,921 It never rolls its eyes on your tenth urge to chase a rally 285 00:10:35,954 --> 00:10:38,041 And it charges no one percent management fee 286 00:10:38,234 --> 00:10:39,473 For ordinary people 287 00:10:39,474 --> 00:10:42,722 This may be the greatest gift AI brings to finance 288 00:10:43,275 --> 00:10:43,622 So 289 00:10:43,795 --> 00:10:45,002 For retail investors 290 00:10:45,035 --> 00:10:47,074 AI is not your competitor 291 00:10:47,074 --> 00:10:48,482 It is your immune system 292 00:10:48,594 --> 00:10:51,034 It protects you not from market risk 293 00:10:51,035 --> 00:10:52,412 But from yourself 294 00:10:52,764 --> 00:10:53,351 The final layer 295 00:10:53,685 --> 00:10:54,612 My judgment 296 00:10:55,004 --> 00:10:58,532 Based on my years observing AI and investing 297 00:10:58,685 --> 00:11:00,211 I believe that over the next five years 298 00:11:00,285 --> 00:11:03,451 Finance will undergo a dramatic polarization 299 00:11:03,565 --> 00:11:05,471 At one extreme is pure quantitative execution 300 00:11:05,805 --> 00:11:07,292 Which AI will take over entirely 301 00:11:07,484 --> 00:11:10,532 Data cleaning financial analysis and risk calculations 302 00:11:10,565 --> 00:11:12,532 Compliance checks and trade execution 303 00:11:12,565 --> 00:11:14,492 These tasks will no longer need people 304 00:11:14,644 --> 00:11:16,032 What do these jobs have in common 305 00:11:16,285 --> 00:11:20,171 Clear rules defined inputs and outputs and standardization 306 00:11:20,325 --> 00:11:22,604 Any job that can be drawn as a flowchart 307 00:11:22,604 --> 00:11:26,234 AI can do faster better and cheaper 308 00:11:26,907 --> 00:11:28,594 At the other extreme is pure judgment 309 00:11:28,787 --> 00:11:30,754 Where human value will actually rise 310 00:11:31,066 --> 00:11:32,494 The greater the uncertainty 311 00:11:32,946 --> 00:11:34,393 And the scarcer the information 312 00:11:34,507 --> 00:11:36,865 And the more a new mental model is required 313 00:11:36,866 --> 00:11:38,074 The more valuable humans become 314 00:11:38,466 --> 00:11:43,433 Early investing crisis decisions cross cultural mergers and disruptive tech pricing 315 00:11:43,507 --> 00:11:45,505 For the foreseeable future 316 00:11:45,507 --> 00:11:46,913 AI still cannot do these things 317 00:11:47,169 --> 00:11:47,653 Why 318 00:11:48,106 --> 00:11:51,153 Because these fields share one trait with no standard answer 319 00:11:51,267 --> 00:11:53,932 Often even the question itself is unclear 320 00:11:54,204 --> 00:11:55,531 That vast middle 321 00:11:55,684 --> 00:11:59,332 The gray zone between execution and judgment 322 00:11:59,365 --> 00:12:01,092 Is the real meat grinder 323 00:12:01,285 --> 00:12:03,083 People who spent ten years as analysts 324 00:12:03,084 --> 00:12:05,891 But never made a major decision independently 325 00:12:06,084 --> 00:12:10,452 People who think mastering a Bloomberg terminal is their core advantage 326 00:12:10,605 --> 00:12:12,771 Their nightmare has already begun 327 00:12:12,964 --> 00:12:13,791 Have you noticed 328 00:12:14,125 --> 00:12:15,092 This structure 329 00:12:15,164 --> 00:12:18,131 Matches the broader impact of AI on labor 330 00:12:18,164 --> 00:12:19,271 Exactly 331 00:12:19,765 --> 00:12:21,072 The bottom tier is not eliminated 332 00:12:21,485 --> 00:12:22,972 It is too cheap 333 00:12:23,204 --> 00:12:24,511 The top tier is not eliminated 334 00:12:24,924 --> 00:12:26,771 AI cannot do top level work 335 00:12:27,005 --> 00:12:29,452 It is always the middle tier that gets eliminated 336 00:12:29,645 --> 00:12:33,131 People who cost too much but lack creativity 337 00:12:33,804 --> 00:12:35,651 If you are in that middle tier 338 00:12:35,885 --> 00:12:38,583 Now is the time to wake up 339 00:12:38,896 --> 00:12:40,402 So let us return to the central question 340 00:12:41,215 --> 00:12:42,015 In the age of AI 341 00:12:42,016 --> 00:12:43,963 Where does the value of human investors lie 342 00:12:44,575 --> 00:12:45,323 I have thought about this for a long time 343 00:12:45,656 --> 00:12:47,463 The answer may surprise you 344 00:12:47,695 --> 00:12:49,983 It is not knowing more than AI 345 00:12:50,175 --> 00:12:53,183 That is already impossible in information processing 346 00:12:53,335 --> 00:12:55,662 Nor is it calculating faster than AI 347 00:12:55,815 --> 00:12:57,183 That is even more impossible 348 00:12:57,295 --> 00:12:59,382 The final value of human investors 349 00:12:59,415 --> 00:13:01,823 Is the willingness to act without knowing enough 350 00:13:01,896 --> 00:13:04,971 And still make a choice with consequences 351 00:13:05,164 --> 00:13:06,471 AI knows everything 352 00:13:06,763 --> 00:13:08,011 But it does not dare to bet 353 00:13:08,203 --> 00:13:09,351 The essence of investing 354 00:13:09,684 --> 00:13:11,323 Is facing uncertainty 355 00:13:11,324 --> 00:13:13,371 And making a costly commitment 356 00:13:13,564 --> 00:13:14,111 That is something 357 00:13:14,363 --> 00:13:16,887 Only a living person can do 358 00:13:17,280 --> 00:13:19,328 But this contains a deeper question 359 00:13:19,560 --> 00:13:20,688 I will not explore it today 360 00:13:20,721 --> 00:13:21,887 I will leave it for you to consider 361 00:13:22,080 --> 00:13:25,408 If AI someday learns to bear consequences 362 00:13:25,481 --> 00:13:27,568 Learns to bet with insufficient information 363 00:13:27,680 --> 00:13:29,487 And learns to build trust with people 364 00:13:29,641 --> 00:13:32,928 Then the line we use to separate humans from machines 365 00:13:33,040 --> 00:13:34,627 Will it disappear completely 366 00:13:35,160 --> 00:13:35,708 Did that line 367 00:13:36,040 --> 00:13:37,147 Ever truly exist 368 00:13:38,173 --> 00:13:39,852 If you work in finance 369 00:13:39,853 --> 00:13:41,500 And feel this pressure 370 00:13:41,612 --> 00:13:43,900 Share your real experience in the comments 371 00:13:44,173 --> 00:13:45,380 In your work 372 00:13:45,532 --> 00:13:47,160 Which tasks has AI already replaced 373 00:13:47,652 --> 00:13:50,753 And which tasks do you think it can never replace 374 00:13:51,369 --> 00:13:51,947 I am Wang Lijie 375 00:13:52,202 --> 00:13:53,074 See you next time