> ## Content Index > Fetch the complete content index at: https://barbarosozturk.com/llms.txt > Use this file to discover other available public pages before exploring further. # AI Will Make It More Difficult to Pick Stocks - URL: https://barbarosozturk.com/blog/ai-will-make-it-more-difficult-to-pick-stocks/ - Published: 2026-09-13T14:46:30.000Z - Updated: 2026-09-13T20:59:17.000Z - Description: The biggest alpha was always finding what the market missed. When everyone has the same AI analyst, nothing is missed — and everything moves together. - Author: Barbaros - Tags: Economy & Markets, Artificial Intelligence The biggest gains are done when people find opportunities that the market misses. That is the entire game of stock picking. Not buying good companies. Buying good companies before the market realizes they are good. Benjamin Graham said markets are a voting machine in the short term, weighing machine in the long term. Your edge was doing the weighing while everyone else was voting. AI just gave everyone the same scale. ![](https://storage.ghost.io/c/64/81/648165b7-979e-43dd-9c43-02101159f467/content/images/2026/09/2-17.jpg) ### 1\. AI Democratizes Research → Markets More Efficient Five years ago, edge = access. You had Bloomberg, expert networks, DCF models you built over weeks, channel checks in Taipei. Now? **1\. AI Democratizes Access to Research & Financial Models** AI tools now open → researchers, startups, students access models & data previously gated. Knowledge for all. Papers, dataset, report, open source financial model — one prompt. Earnings call transcript summarized in 2 seconds, 10-K risks extracted, supplier chain mapped, competitor positioning charted. **2\. Markets More Efficient** Real-time analysis, broader data → faster pricing, less inefficiency. Markets reflect info now. Price = Information. Data flow instant. When everyone has the same Muse, Claude, Gemini analyst reading the same filings at the same speed, mispricings close in minutes, not months. The market becomes brutally efficient at pricing known information. **Result:** Less inefficiency to exploit. Traditional alpha from "I read the footnote on page 147" disappears. ![](https://storage.ghost.io/c/64/81/648165b7-979e-43dd-9c43-02101159f467/content/images/2026/09/3-21.jpg) ### 2\. Groupthink Voting Machine — Short Term Amplified Here is the paradox. More efficiency should make markets rational. Instead, AI amplifies groupthink. **3\. Groupthink Voting Machine — Short-Term** Consensus bias, herd behavior. Quick unanimous votes, short-term focus, ignores diverse signals. FOLLOW THE CROWD. SAME! SAME! Why? Everyone prompts similar models trained on same data with same objective: "Summarize sentiment, give price target." If 10,000 funds ask ChatGPT "Is Google winning AI race?" in Q1 2023, they get similar answer: "Google caught flat-footed, slow, bureaucratic." So everyone votes SELL. Price drops not on fundamentals but on consensus narrative generated by same AI. Markets are a voting machine in the short term. AI makes the voting faster, louder, and more correlated. It is 10,000 voters all reading same AI-generated voting pamphlet. This amplifies momentum and groupthink. Stocks move together more in short term because models herd. ![](https://storage.ghost.io/c/64/81/648165b7-979e-43dd-9c43-02101159f467/content/images/2026/09/4-11.jpg) ### 3\. Opposite Opportunities: Google Undermined Case **4\. Google Undermined in AI Race → Opposite Opportunity** Incumbent, slow, bureaucratic → Undermined → Opportunity for others. Open AI ecosystem, new entrants, innovation. Legacy player held back; gap opens for nimble players to lead. Perfect example: 2023-2024\. Consensus: Google is done. OpenAI won. Bard is joke. AI kills search. Groupthink voting machine pushed GOOGL to 95, PE 18, everyone short innovation. Opposite opportunity: Those who did weighing, not voting, realized Google had DeepMind since 2014, owned Transformer paper, owned data centers, owned distribution. The AI race narrative was voting, not weighing. Those who bought the dip won big. AI creates more of these moments. When everyone has same research, contrarian signal gets drowned. The crowd all thinks Google is slow, same! same! — but nimble players, open AI ecosystem, new entrants actually get fuel from incumbent being undermined. This is the new inefficiency: not information asymmetry, but narrative asymmetry. ![](https://storage.ghost.io/c/64/81/648165b7-979e-43dd-9c43-02101159f467/content/images/2026/09/image_1.jpg) ### Conclusion: Secret Winners Win Big, But Harder to Find Secret Winners Win Big Against Consensus Amplified Groupthink — The Consensus: Everyone says SELL! Follow the herd! It's obvious! Just go with the crowd! No risk, just consensus! Group Trend down. Secret Winners — The Contrarian Bet: Small Bet → WIN BIG +500%+ Trophy chest overflowing. Contrarian Bet: BUY LOW, Unpopular Pick. Difficult to find — few choose the opposite path. When they are right, payoff is outsized. Before AI: 100 secret winners a year. You needed to be smart to find 10. After AI: 10 secret winners a year because markets price 90 instantly. You need to be contrarian, patient, and willing to look wrong for quarters while voting machine screams SAME! So it will be more difficult to find secret winners who win against the consensus but those who win will win big. Because when everyone is democratized, the only edge left is behavioral — being willing to vote differently when AI tells everyone to vote the same. Edge shifts from research to temperament. From "I know more" to "I can endure being alone longer than the model herd." That is harder. And more lucrative.