80-Year-Old Investment Master Howard Marks Deep Interview: Why Did I Change My View on AI? Second-Level Thinking and 25 Years of Decision Reflection

This episode with Howard Marks is one of the most valuable in terms of investment cognition. Because Marks is not teaching you "which stock to buy," nor is he trying to prove how accurate his predictions of the future are. What he is really discussing are the three most difficult aspects of professional investing: Making decisions without knowing the future; Forming your own judgments outside of market consensus; Even when forming judgments, always leaving room for "I could be wrong." These three aspects combined represent the core capability of institutional investing: Decision-Making Under Uncertainty. Moreover, the overall quality of your summary is very high, but I suggest upgrading a few numbers and understandings first. ──────────────── 1. First, correct the most important point: "Wasted 25 years" is actually a title packaging. Howard Marks' original intention in this My First Million interview was not: "I completely wasted the first 25 years of my life." What he said was that looking back at his career choices from entering Citibank in 1969 to co-founding Oaktree with Bruce Karsh in 1995, many were not made after high awareness and systematic planning. For example: Why go to Citibank's research department? Because he had a good experience working there the previous summer. Why later switch to bonds? Because he did poorly in stock research and was reassigned. Why move to California in 1980? He even described the randomness of it as "sunshine and palm trees." Why eventually enter High Yield? Because in 1978, a boss suddenly asked him to research "a guy named Milken and something called high-yield bonds." Marks himself referred to these as paths formed by many coincidences, passivity, and luck. So a more accurate title should be: "I only truly began to consciously lead my life around the age of 50." And not: "The first 25 years were meaningless." In fact, those 26 years accumulated the core experiences of Oaktree's later High Yield, Distressed Debt, and cyclical knowledge. Oaktree's official biography shows that he was at Citicorp from 1969 to 1985, at TCW from 1985 to 1995, and only then co-founded Oaktree. ──────────────── 2. This reveals a very advanced life principle: Path Dependence. The most interesting part of Marks' story is: The path to ultimate success is not necessarily pre-designed. If that phone call in 1978 had not been made; If he had gone out for lunch that day; If he had been very successful in stock research and not been reassigned; He might not have become the Howard Marks of today. This is called: Path Dependence. Life is not: A → B → C → D unfolding as planned. More often it is: A → Random Event → New Ability → New Connections → New Opportunities → New Path. So the real importance of life planning is not: Predicting your career at 50 accurately at 20. This is almost impossible. But rather: Continuously increasing the number and quality of future paths available. This is Optionality. ──────────────── 3. Therefore, what young people should really optimize is not a "perfect plan," but "high optionality." For example, in your 20s: Learn programming; Study finance; Practice sales; Learn English; Build industry networks; Understand AI; Accumulate cash. What do these skills have in common? They do not correspond to just one job. They open up: Many future paths. Conversely, a skill that is highly narrow, non-transferable, and tied to a declining industry: Has low optionality. So a very important principle of capital allocation in youth is: Prioritize investing in skills that can expand future choices. You cannot predict the future. But you can ensure that when opportunities arise in the future: You are qualified to seize them. ──────────────── 4. The real strength of Howard Marks has never been predicting the future. Oaktree's own investment philosophy even explicitly states: Macro-forecasting not critical to investing and: Disavowal of market timing. This is very counterintuitive. Because financial media asks every day: When will the Fed cut rates? What will next year's GDP be? What will the S&P 500 be at the end of the year? When will the AI bubble burst? Howard Marks' core attitude is actually: Many important future variables are essentially unknowable. But: Unpredictable ≠ Uninvestable. This is the fundamental difference between him and many financial commentators. ──────────────── 5. The real investment question is not "What will happen in the future?" but "What does the current price imply about the future?" This is a threshold that must be crossed to understand Howard Marks. Ordinary investors ask: "Is the company good?" Howard asks: "Given how good the company is, how much has that been reflected in the price?" For example, a world-class AI company: Future revenue growth of 30%. Very good. But suppose the stock price has already been priced based on: Future growth of 50%. Then: The company can be great, But the stock can be poor. Conversely: An ordinary company. The market thinks it will go bankrupt soon. Bond price: 40 cents on the dollar. The final liquidation can recover 70. The company is poor. But the investment could be very good. So: Good Asset ≠ Good Investment. What truly determines investment returns is: Price relative to reality. ──────────────── 6. This is the true meaning of "Second-Level Thinking." Marks directly states in the interview: If what you see is exactly the same as everyone else, you cannot achieve superior performance. To surpass the average: Your judgment must be: Different And: Better. Note the second word. This is very important. Many people misunderstand Second-Level Thinking as: "If others are bullish, I will be bearish." Wrong. That is: Contrarian for the sake of being contrarian. ──────────────── 7. True Second-Level Thinking is "Consensus + Price + My Different Judgment." For example. The market unanimously believes: A certain AI company will be great in the future. First-level thinking: AI is great, so buy. Second-level thinking: Everyone thinks AI is great, so has the current price already reflected extreme optimism? Continuing: The market expects revenue to grow 40% in the future. I believe: Only 25%. So even if the company grows 25%, which is very good, The stock may still fall. Conversely: The market believes a certain company: Is doomed. I believe: It is just difficult, it won't die. Then a "poor company" may become: An excellent security. This is: Variant Perception. ──────────────── 8. Therefore, excellent investing truly needs to satisfy a brutal formula. Different + Correct. Only Different: There are many crazies. Only Correct: But the market already knows, There is no Alpha. You must: Be different And: Ultimately, the facts prove you are closer to correct. This is why it is extremely difficult to consistently outperform the market. Because you must find: Consensus Error. ──────────────── 9. Howard says "Second-Level Thinking cannot be taught," I believe it should be understood in two parts. He makes it very clear: He can tell you: Second-Level Thinking is important. But cannot directly teach someone: How to form an insight that is different from the market and also correct. He even uses basketball as an analogy: Some things are like height, not trained. I only agree with this point halfway. ──────────────── 10. "Genius insight" may not be replicable, but the second-level thinking process can be fully trained. For example, when analyzing any investment, force yourself to answer: What is the market consensus? What assumptions does the current price imply? Which assumptions are the most fragile? If the consensus is correct, how much upside is there? If the consensus is wrong, how much downside is there? What information is the market already aware of? What do I know that the market does not? Why might I be right? What evidence would emerge that I must acknowledge I was wrong? This entire set can be trained. You cannot train to become: Howard Marks. But you can train to: Make half as many foolish mistakes as you did yesterday. This in itself is of great value. ──────────────── 11. The AI part is particularly brilliant because Howard did something very rare: publicly modifying his worldview. In December 2025, he published "Is It a Bubble?" focusing on analyzing AI from the perspective of historical bubbles: the technology may be real, but the investment enthusiasm could still be excessive; great technology and poor investment can coexist. Just about 11 weeks later, he wrote "AI Hurtles Ahead." The reason is simple: His son Andrew told him: AI is developing too fast, and your piece needs to be updated. Thus, Howard basically rethought everything. This is very much Howard Marks. Because true intelligence is not: Always proving yesterday's self correct. But: Modifying models when new facts emerge. ──────────────── 12. This is called Bayesian Updating. Assuming initially you believe: The probability of AI becoming a super significant technological revolution is: 60%. Then you receive: Agent capabilities improve; Autonomously writing code; Enterprise adoption; Model reasoning enhancement; Cost reduction and other new evidence. A reasonable person should update the probability: 60% → 75% → 85%. And not because: "I said last year that AI is a bubble" So always insist it is a bubble. A truly mature investor has no: Identity Attachment. They do not turn: Their own views Into: Their identity. ──────────────── 13. Howard now believes the truly special aspect of AI is Autonomy. He said in the past: Railroads; Computers; The internet Were essentially: Tools. Humans decide: What to do; How to do it. AI introduces a very important change: You tell it the goal, and it can partially find ways to achieve that goal by itself. This is Autonomy. This distinction is very important. If the previous generation of software was: Tool Software Then the agent era is evolving towards: Labor Software. In the past, software: Helped employees work. In the future, some AIs: Will directly complete parts of employees' entire workflows. The economic implications are completely different. ──────────────── 14. This is also why Howard is increasingly worried about AI's impact on the labor market. His formal memo for 2026 has already clearly expressed: AI may replace a large number of "thinking jobs," while machines controlled by AI may also replace some "doing jobs." He even raised the concepts of: joblessness and: purposelessness. The second term is deeper than the first. The long-term issues brought by AI may not only be: "Humans have no wages." But rather: "If machines can do everything better than I can, what is my social significance?" This has transformed from an economic issue into: A philosophical issue. ──────────────── 15. But Howard did not jump from "AI is great" to "buy AI stocks with closed eyes." This is the most important investment discipline of this entire segment. His latest conclusion is very clear: AI is very real. The potential of AI may even still be underestimated. However: This does not mean that AI-related asset prices are cheap. He even emphasizes: Technology potential and: Investment price are two different things. This may be a statement that all AI investors should write on the wall. ──────────────── 16. Great technological revolutions in history can completely lead investors to suffer significant losses. Railroads changed the world. Many railroad investors still went bankrupt. The internet changed the world. In 2000, many dot-com stocks went to zero. Fiber optics eventually became the digital economy's infrastructure. But in the late 1990s, there was serious overbuilding. So: Technological Truth ≠ Investment Truth. The technological judgment: "AI will change the world." May be 100% correct. The investment judgment: "Any AI company is worth this price now." May be completely wrong. ──────────────── 17. The real investment question should be broken down into four layers. First: Technology Is the technology real? Second: Adoption Does the market really need it? Third: Economics Who will ultimately make money? Fourth: Valuation Does today's price reflect or even overextend all three of the above? Many people only complete the first layer: "AI is great." Then directly: Buy. Professional investors must complete all four layers. ──────────────── 18. Howard's final attitude towards AI investment is actually extremely classic. He did not say: All-in. Nor did he say: All-out. His 2026 memo continues to maintain his conclusion: You cannot completely miss what may be a historically significant technological revolution due to bubble risks; Nor can you take on ruin risk just because the technology is great. Therefore: Moderate position + Selectivity + Prudence. Translated into investment language, this means: Position Sizing. Many people think the most important thing in investing is: What to buy. In fact, in professional investing: How much to buy is equally important. ──────────────── 19. This connects to Howard's most important philosophy: "I could be wrong." He said a very memorable phrase: Starting a sentence with: "I could be wrong, but..." Or: "I don't know, but..." Rarely leads to huge disasters. The real danger is: "I'm 100% convinced..." Why? Because being wrong in itself will not bankrupt you. What truly leads to bankruptcy is: Wrong view × Huge position. ──────────────── 20. This is one of the most important distinctions in the investment world: Probability and Position are two different things. Assuming you believe a certain event: Has an 80% success rate. Many people translate this in their minds as: "Almost certain to succeed." So: All-in. But mathematically: A 20% chance of failure Means: In five possible worlds, About one will fail. If failing once means going to zero: Then this bet may not be worth making at all. So a truly professional investor does not ask: "How confident am I?" But rather: "Even if I'm wrong, can I still play?" ──────────────── 21. This is called Ruin Constraint. Before any investment question, first ask: Can I survive being wrong? If the answer is: No. Then even if the expected return is high, it may not be worth it. This is why: Buffett; Marks; Taleb Although their styles are completely different, they all share a common focus: Avoid permanent ruin. Only by surviving: Can compounding continue. ──────────────── 22. Therefore, the true compounding formula should add a condition at the front. People like to write: Wealth = Capital × (1+r)^t But the real version should be: Wealth = Survival × Capital × Compounding. If Survival = 0, Then everything else: Is 0. This is why Howard has long emphasized Risk Control, and Oaktree lists "Primacy of Risk Control" as the first principle of the entire company's investment philosophy. ──────────────── 23. The story from 2008 is the perfect practical embodiment of all of Howard's philosophies. In 2007-2008, Oaktree raised about: $11 billion for distressed debt investments. At that time, Oaktree's largest similar fund was only about $2.5 billion, so this scale was extremely aggressive. Howard himself clearly stated this number in the program. Why were they able to raise so much? Not because: They suddenly produced a very nice Pitch Deck. But because they had been doing distressed debt for about 20 years; Had experienced 1990-91; 1998; 2001-02; Had long-term performance; Had LP relationships; Had credibility. Howard himself referred to this as: Reservoir of goodwill—long-term accumulated credibility. ──────────────── 24. So the "ark must be built before the flood" actually speaks to the liquidity paradox of capital markets. When you need money the most: It is the hardest to raise money. In a booming market: LPs say: Of course, we will give you money. In a market crash: LPs panic themselves; Assets shrink; Liquidity tightens; Committees are afraid to approve. As a result: At the time of the greatest opportunity, capital is the most scarce. So truly excellent contrarian investment institutions must achieve: Capital before opportunity. This is what is called: Dry Powder. ──────────────── 25. This is also why excellent capital allocators often appear "foolish" during bull markets. When the bull market is at its craziest: Others: Leverage up; Chase hot stocks; Increase valuations; Expand funds. Howard instead: Increases defenses; Raises crisis funds; Retains capital. It may seem: Returns lag. But the real goal is not: To win every quarter. But rather: To still have money when extreme mispricing occurs. This is a completely different game. ──────────────── 26. The real numbers worth studying from 2008 also need slight correction. In the program, Howard said: Bruce Karsh, in the last 15 weeks of 2008, Averaged: About $450 million per week Invested about: $7 billion. Howard's latest memo for 2026 also confirms again: Bruce averaged about $450 million per week for 15 weeks. But another official historical review from Oaktree gives a more precise figure: From September 18, 2008, to the end of the year, Bruce's team averaged about $400 million per week, totaling about $6 billion; Plus purchases from other Oaktree teams, totaling: About $7.5 billion. So the different figures of: $6 billion; $7 billion; $7.5 billion are not necessarily contradictory. They correspond to: Different teams and statistical measures. The most prudent description is: From Lehman's collapse to the end of 2008, Oaktree invested about $7 billion in capital over approximately one quarter. ──────────────── 27. But the most impressive part is not "daring to bottom fish." Many people also said in 2008: It's cheap. The real difficulty is: Deploying Size. Saying: "This bond is cheap." Is easy. Actually buying: $450 million in a week. Continuing in the second week. Continuing in the third week. The market continues to fall. News gets worse. Clients are scared. Colleagues are doubtful. And then continuing to buy. This is where institutional investing becomes truly terrifying. ──────────────── 28. Howard and Bruce even constantly doubted whether they were buying too fast or too slow. Howard later recalled: Bruce sometimes ran over saying: Are we buying too slowly? The next day he might say: Are we buying too fast? But they still continued to invest. This point is very important. Because ordinary people often think: When masters make great investments, Their inner state must be: 100% Conviction. Wrong. Many great investments are in a state of: High uncertainty + Still acting. ──────────────── 29. This is called Investing with Trepidation. Market lows will never send you an email saying: "Today is the bottom, feel free to buy." The real bottom usually comes with: Bankruptcies; Layoffs; Banking crises; Wars; Pandemics; Liquidity crises. If all fear disappears: Assets usually have already risen back. Howard's meaning in the interview is very clear: Waiting until there is nothing to fear, the opportunity has often passed. ──────────────── 30. However, "buying in fear" does not mean "buying just because prices are down." This is where many people easily misunderstand Howard Marks. In 2008, Oaktree did not: "Buy with closed eyes just because prices fell a lot." Howard said that at that time they could buy some debt instruments, even if the related PE acquisition companies ultimately fell to a quarter or even one-fifth of their original acquisition value, the debt investments could still have strong protection. In other words: There is: Margin of Safety. The real contrarian investment formula is: Panic + Mispricing + Fundamental Protection And not: Price Down = Buy. ──────────────── 31. This is also why Distressed Debt is particularly suitable for Howard Marks. Stocks: The company is worth 100. You buy at 80. The downside may be: 0. Debt instruments: If you are higher up in the capital structure, And the company's asset value is enough to cover your claim, It may be: The company does not need to return to its former glory, And you can still make money. So the core issue with Distressed Debt is often not: Can the company become great again? But rather: What is the Recovery Value? This is Credit Thinking. ──────────────── 32. Howard's world is completely different from that of many growth stock investors. Growth Investors ask: How good can the future be? Credit Investors first ask: If it goes poorly, how much can I get back? The former emphasizes: Upside. The latter emphasizes: Downside. Howard's entire thought system has long been influenced by credit investment training. So his first reaction is never: "How much can I earn?" But rather: "How do I lose? How much will I lose?" ──────────────── 33. This also explains why he particularly admires "Fooled by Randomness." One of Taleb's most important contributions is reminding investors: Outcome cannot be retroactively attributed to Skill. A person this year: +80%. Does not mean: Genius. It could be: High leverage; Single position; Good luck. Another person: +8%. May instead be: Very low risk; Excellent process; Long-term stability. Howard recommends "Fooled by Randomness" in the program, clearly emphasizing the significant role of randomness in short-term results and that one cannot simply infer investment ability from attractive performance. ──────────────── 34. This raises an extremely important question in the investment industry: Skill or Luck? Assuming a fund manager: Has outperformed for three consecutive years. Skill? Possibly. If 50 people are flipping coins: Someone will inevitably get heads five times in a row. This is why institutional investors truly need to study: Investment process; Risk-taking; Position; Beta; Leverage; Cycles; Drawdowns; Repeatability. And cannot just look at: Annual Return. ──────────────── 35. Why is "A Short History of Financial Euphoria" another of Howard's favorite books? Because Galbraith does not study: A specific valuation model. But rather: Human nature. Technology has always changed. Financial products have always changed. But: Greed; FOMO; Leverage; "This time is different"; Taking price increases as proof of ability; Has not significantly upgraded over hundreds of years. Howard himself said in the interview that this book influenced his understanding of the psychological weaknesses behind Boom/Bust cycles. ──────────────── 36. Therefore, the real cycle is not economic indicators, but human psychology. A complete cycle often looks like this: Pessimism. Assets are cheap. Smart money buys. Economic improvement. Prices rise. The public believes. Leverage increases. Risk awareness decreases. "New era." Over-financing. Over-investment. Disappointment. Prices fall. Panic. Forced selling. And then: Restart. So Howard's concept of Market Cycle is essentially a large part of: Psychology Cycle. ──────────────── 37. Why does he not believe in accurately predicting market tops and bottoms? Because you can know: The market is overheated. But you do not know: How much longer it can stay hot. In 1998, you thought tech stocks were expensive. You may be correct. But the Nasdaq could continue to rise for another two years. In 2006, you thought real estate was dangerous. You may be correct. But the crisis may take time to truly erupt. So Howard values: Temperature More than: Clock. Not predicting: "The peak will be on October 17." But rather judging: The current market is: Cold; Normal; Hot; Extremely hot. And then adjusting: Aggressiveness. This is much more realistic than predicting specific dates. Oaktree has long emphasized adjusting defenses and offensives based on market conditions rather than precise Market Timing. ──────────────── 38. Looking again at his 39-year relationship with Bruce Karsh, this part of the business value is severely underestimated. As of this 2026 interview, the two have collaborated for about: 39 years. Howard said: In 39 years, there have been many knowledge differences, But no real fights. The foundation is: Mutual Respect. He gives a very good formula: Shared Values + Complementary Skills. Shared values. Complementary abilities. ──────────────── 39. Why are Shared Values more important than Shared Skills? Assuming two partners are very smart. But: A is willing to leverage heavily. B is extremely conservative. In a bull market: A thinks B is holding back. In a bear market: B thinks A is going to kill the company. For example: A believes in never deceiving clients. B believes in doing whatever is legal. Ultimately, there will be conflict. So: Strategy can differ, but Values cannot conflict. The biggest bomb in a long-term relationship is often not: Poor ability. But rather: Different value functions. ──────────────── 40. Why are Complementary Skills also very important? Howard himself puts it very plainly: Bruce: Does investments; Does distressed; Strong in legal and restructuring; More willing to stay behind the scenes managing funds. Howard: Clients; Fundraising; Communication; Memo writing; External expression. The two do not: Duplicate each other. But rather: 1 + 1 > 2. Howard even says: A truly good partner is someone who is willing and good at doing: What you cannot or do not want to do. ──────────────── 41. Therefore, many startup teams' biggest mistake is: looking for "another self." Two Product Guys. No one sells. Two Sales Guys. No one makes products. Two Visionaries. No one operates. Two Alpha CEOs. Every day fighting for control. A truly excellent co-founder combination should pursue: Values overlap, skills don't fully overlap. High overlap in values. Skills should be as complementary as possible. This is much more important than: "We are best friends." ──────────────── 42. Howard also added a third condition that I think is very advanced: Appreciation. This point is not mentioned in many business books. After cooperating for a long time, People easily start to feel: "I contribute the most." Then: The other party's income is too high; Equity too much; Fame too great. Howard says that a truly long-term partnership must also: Appreciate the other party for doing what you are unwilling or unable to do. So the complete formula for long-term partnerships can actually be written as: Shared Values × Complementary Skills × Mutual Respect × Appreciation. Any one of these long-term factors going to zero, The cooperation may end. ──────────────── 43. This is why he immediately thinks of Buffett and Munger. Howard has indeed had long interactions with Buffett. He told a great story in the program: Oaktree was a major creditor in the Enron-related Osprey debt restructuring; Berkshire was also a significant creditor; Buffett entrusted the proxy to Bruce Karsh to handle. Later, Buffett was very satisfied with Bruce's handling, and they established a relationship. Interestingly, Howard said that the reason he later wrote "The Most Important Thing" was largely because Buffett encouraged him: You should write a book. ──────────────── 44. Your summary of Munger's greatest investment contribution to Buffett is basically correct. In his early years, Buffett was influenced by Benjamin Graham and liked: Cigar Butt Investing. That is: Extremely cheap; Average quality; Companies with "last few puffs of value." Munger later pushed him to expand the definition of "Value": No longer just pursuing: Low-priced ordinary companies. But starting to accept: Buying excellent companies at reasonable prices. Howard himself also clearly summarizes Buffett and Munger's transformation in Oaktree's official memos. ──────────────── 45. Why is this transformation so important? Because as capital scales up: Cigar Butt is hard to scale. Assuming you manage: $1 million. You find a small company worth only $5 million that is severely undervalued. Great. You can make a lot. If you manage: $500 billion. Even if this opportunity doubles: It is almost meaningless to the overall portfolio. So the larger the capital scale, The more it needs: Scalable Quality. This is also why Buffett later became willing to hold: Coca-Cola; GEICO; Apple And other huge companies for the long term. ──────────────── 46. There is actually a very important evolution in investing. Early value investing: Price first. Looking for extremely low prices. Mature value investing: Business quality + Price. Because truly excellent companies can: Reinvest; Compound; Expand their Moat; Generate cash over the long term. Thus, time itself begins to work for you. This is the distinction between: Cheapness and: Compounding Quality. ──────────────── 47. What is most admirable about Howard discussing AI at 80 is not that he understands AI. Rather, it is: He openly states he is not an expert. Oaktree's AI Memo begins by clearly acknowledging that he is not a techie and has no special expertise in AI. However: He still learns. Asks his son. Questions models. Modifies his views. Rewrites memos. This indicates that true advanced intelligence is: Strong opinions, loosely held. You can have clear judgments. But cannot marry your judgments. ──────────────── 48. This may also be the most important ability of humans in the AI era: Model Updating. The half-life of knowledge is getting shorter. The tools; Models; Industry rules; Technical abilities you learn today may become outdated in three years. Therefore, in the past, society rewarded: Knowing more. In the future, it will increasingly reward: Updating faster. If an 80-year-old investor can rewrite an entire memo because of a single sentence from his son: "Dad, the world has changed." This in itself is a strong competitive advantage. ──────────────── 49. However, I believe Howard's judgment that "AI cannot identify character" cannot be concluded yet. He gave a very good investment example in the interview: Sometimes when meeting a manager, The data is fine; The business plan is also fine; But you just feel: A chill down your neck. Thinking: This person is not right. Howard believes this kind of experiential judgment may still belong to human advantages. This view is reasonable now. But it cannot further assert: AI will never be able to do this. Because in the future, AI may analyze: Inconsistencies in language; Historical behaviors; Legal records; Accounting anomalies; Voice; Communication patterns; Millions of historical fraud cases. It may even discover: Patterns that human intuition cannot articulate. ──────────────── 50. Therefore, the true human moat may not be "intuition," but responsibility. This is where I believe I can push further than Howard. In the future, AI may: Analyze better than humans; Read materials faster than humans; Discover more patterns than humans. But in the end, investing must still have someone deciding: To bet $1 billion or not? Because: AI has no clients; No LPs; No career; No legal responsibilities; No pain of capital loss. So the core role of humans may shift from: Information Processor To: Judgment + Accountability. AI provides: Probability distributions. Humans decide: Positions. This may be the strongest combination in the future. ──────────────── 51. Howard's true investment advantage may never have been "knowing more facts." He reads: The same newspapers as everyone else. Watches: The same news as everyone else. In 2008: He saw the bad news just like everyone else. The real difference is: Others see: The world is terrifying → Sell. He sees: The world is terrifying → How much has this terror been reflected in the price? This is a completely different cognitive loop. So investment Alpha often does not come from: Information Advantage. But from: Interpretation Advantage. The same set of facts. Different conclusions. ──────────────── 52. This is also where AI may truly change the investment industry. Part of the value of analysts in the past was: Finding information; Organizing financial reports; Building models; Reading announcements; Comparing companies. AI will turn a lot of basic analysis into: Commodity. Thus, what remains in the investment industry that is increasingly valuable includes: Defining problems; Variant Perception; Position Sizing; Risk Management; Judging management; Understanding market psychology; Executing in panic. In other words: AI may first eliminate "information-moving investors," rather than top capital allocators. Howard himself also judges in the interview that AI may expose a batch of investment managers who were originally overvalued, just as indexing once exposed many active managers who could not consistently outperform. ──────────────── 53. What Howard means by "the only success is living life your own way" is actually more important than investment philosophy. He quotes Christopher Morley: The true success in life is: Live your life your own way. Then his advice is very specific: Find something: That plays to your strengths; Avoids your weaknesses; And makes you happy. And you cannot let: Friends; Parents; Society decide for you just because they think a certain path is good. This is essentially about: Personal Capital Allocation. ──────────────── 54. A person is also a portfolio. You have: Time; Health; Attention; Talent; Relationships; Money. These are all limited capital. The real question in life is: Where to allocate this capital? If a person: Is extremely good at writing; Likes independent thinking; Hates managing 500 people, And ends up forcing themselves to be a large company CEO for the sake of "success definition," It may be a very poor capital allocation. Another person: Is good at sales; Likes competition; Likes people; But hides behind doing analysis, Also wastes their talent. ──────────────── 55. Therefore, "being oneself" is definitely not "doing whatever you want." The advanced version should be: Find comparative advantage in yourself. What am I stronger at than others? What can I persist in long-term? What activities give me energy rather than consume it? What will the market pay for? The intersection of these four questions may form: Career Alpha. This is similar to investing: It is not about finding: The "best" profession in the world. But about finding: Opportunities where price and value are most mismatched for you. ──────────────── 56. Howard founded Oaktree at 50, providing a very interesting counterexample to age anxiety. Oaktree was established in: 1995. Howard was about: 49 years old at the time. Today, many startup narratives like: 19 years old; 22 years old; Dropping out; Unicorns. But when Howard truly turned his name into a global institutional investment brand, He was already nearing 50. Why? Because industries like Credit and Distressed truly require: Cyclical experience; Judgment; LP trust; Relationships; Reputation; Capital. These things: Require time to compound. So the founder peak varies greatly across different industries. ──────────────── 57. More interestingly, Oaktree itself has entered a new capital cycle. In 2019, Brookfield acquired a majority stake in Oaktree; by August 3, 2026, Brookfield completed the full acquisition and further integration of Oaktree. Oaktree's official now directly describes 2026 as a stage of comprehensive integration of the two major platforms. This is very interesting. Howard at 80 discusses: Life; Partnerships; Cycles; AI. Meanwhile: The company he founded 30 years ago is entering: An institutional lifecycle longer than the founder himself. This is what is called: Institutionalization. Truly great companies must ultimately: Outlive the founder. ──────────────── 58. Howard's life perfectly embodies the joint effect of "Luck + Skill." If you only listen to successful people say: "I just have good judgment." It is easy to generate: Narrative Fallacy. Howard instead actively states: Many things are just luck. Being reassigned to the bond department. Encountering High Yield. Meeting Milken's era. Meeting Bruce Karsh. But luck is not everything. Because after opportunities arise: He spent decades building: Knowledge; Systems; Clients; Brands; Risk control; Institutions. So the true success formula should be: Luck creates exposure. Skill converts exposure into compounding. Luck brings opportunities to you. Skill determines: Whether you can keep them. ──────────────── 59. If I were to compress the entire episode with Howard Marks into eight actionable principles: First: Do not predict the future; study what the current price reflects about the future. ──────────────── Second: Different does not have value; "different and correct" has Alpha. ──────────────── Third: Always write "I could be wrong" into position sizing, not just hang it on your lips. ──────────────── Fourth: Do not wait for fear to disappear before investing, as cheap prices often coincide with fear. ──────────────── Fifth: Contrarian investing is not buying just because prices fall, but because panic has created verifiable mispricing. ──────────────── Sixth: Truly long-term partnerships require shared values, complementary skills, respect, and appreciation. ──────────────── Seventh: Modifying views after new facts emerge is not weakness, but intellectual discipline. ──────────────── Eighth: Life and investing are essentially both capital allocation. You only have: Limited time. So do not let: Friends; Society; Parents; Inertia allocate for you. ──────────────── 60. If I were to retitle this episode, I would elevate it from the original "Wasted 25 Years" to: "Howard Marks' Ultimate Investment Lesson After 80: True Masters Do Not Predict the Future, But Manage 'I Might Be Wrong.'" Subtitle: From the $7 billion contrarian build-up in 2008 to a renewed understanding of AI: Second-Level Thinking, probability, cycles, partners, and life choices after 50. Because this is the true essence of the entire episode. ──────────────── 61. Finally, I believe Howard Marks' highest level is not that he can see what others cannot see, But that he possesses two seemingly contradictory abilities: Sufficient confidence to dare to be different from the market; Sufficient humility to know he might be wrong. Most people can only possess one. Overconfident people: Dare to bet, But easily self-destruct. Overly doubtful people: Are very cautious, But never take action. Truly excellent investors must be in the middle: Conviction without Certainty. Have judgments, But no illusions. Have positions, But do not bet their lives. Have views, But allow facts to overturn them. Have fears, Yet still execute. This is why in 2008, what was truly valuable was not that Howard Marks "predicted the financial crisis." In fact, he constantly emphasizes: Nobody Knows. What was truly valuable was: When everyone did not know the future, he and Bruce Karsh still established a system capable of making decisions, controlling losses, and continuing to bet. This is actually much more advanced than "accurately predicting the next crisis." Because predictions can only be used once. A Decision System can be used for a lifetime.