Recording intelligenceAI-generated brief · check the source for context

Howard Marks on AI and Human Judgment

46:30 recording · EN · 2 speakers

Listen to the original

Listen to the episode
Executive Summary AI
  • Howard Marks explains he rewrote his AI memo entirely after his son Andrew, a VC who deals with AI every day, told him in early February that so much had happened since the first memo around December 9th.0:50
  • He upgraded his view because AI has two qualities no earlier technology had — autonomy, the ability to be given a job without being told how to do it, and an unpredictability that makes the shape of the future unknowable, unlike the railroad, computers or the internet.2:23
  • On conviction, he recounts Lehman going under on September 15th of 08, when Oaktree had raised an 11 billion dollar distressed debt fund in 07-08 — against the 2.5 billion dollar 02 fund that was the biggest in history before — and Bruce Karsh deployed an average of $450 million a week for 15 weeks, $7 billion in a quarter, with no data and no precedent, only supposition.10:38
  • He credits a 39-year partnership with Bruce Karsh, built on mutual respect, shared values and complementary skills, and says they have had intellectual disagreements but never a fight, partly because neither is a financial maximizer.20:58
  • On careers he admits he made his own early decisions haphazardly until he left to start Oaktree around 95, and now tells students to find work that plays to their strengths, avoids their weaknesses and makes them happy, quoting Christopher Morley that there is only one success — to live your life your own way.28:43

Brief overview

Howard Marks says AI's autonomy is unprecedented, but judgment, partnership and second level thinking still decide investing.

  1. Autonomy is what makes AI differentEvery prior innovation was a tool that sped things up; AI can be given a job and figure out how to do it itself.
  2. Big calls are made with doubt, not convictionMarks invested through Lehman while absolutely not confident — a battle hero is afraid and does it anyway.
  3. Partnerships need shared values and complementary skillsCowboys and chickens disparage each other in both good and bad times; Marks and Karsh have lasted 39 years on mutual respect.

Questions this recording answers

10 questions, each answered where it is said
Why did Howard Marks change his mind about AI?

The story is simple: his son Andrew, a VC dealing with AI every day, whose companies use and some create AI, told him in early February that so much had happened since the first memo, written around December 9th, that he had to update it. Marks rewrote the memo entirely.

Answered around 0:50
What makes AI unprecedented compared with earlier technologies?

Autonomy is the obvious one. Every other innovation, from the railroad to computers to the internet, was a tool to speed things up and increase productivity. You can give AI a job without telling it how and it figures it out. The other quality is being unpredictable: nobody knows the shape of the future.

Answered around 2:23
Is there anything in investing that AI cannot do?

There will always be things for which there is no history to train on, and a big percentage of what AI does is knowing history and extrapolating patterns. Some people better understand the probability distribution defining future events. Marks also credits avoiding bad people, when the hair on your neck goes up.

Answered around 6:27
What is second level thinking?

Second level thinking says that if you don't see anything different from everybody else, you can't possibly be superior. You need a variant perception: that the consensus overstates the company's quality, growth rate, earning power, or the multiple it deserves. Then you have to bet on your perception, and you have to be right.

Answered around 8:29
Can someone be taught to be a second level thinker?

Marks says the answer is no. In the book he writes that he doesn't know, but thinks it's more no than yes. He can teach the importance of being a second level thinker, as that chapter does, but not how to hold perceptions at odds with the consensus and be correct. Like height, insight can't be coached.

Answered around 9:13
How did Oaktree decide to invest when Lehman went under?

They had raised an $11 billion distressed debt fund in 07-08, against the prior record, their $2.5 billion 02 fund. At the Lehman bankruptcy they had no data and no prior experience, only supposition. They reasoned that if the financial world melted down it wouldn't matter, so Bruce invested about $450 million a week for fifteen weeks.

Answered around 10:38
What does it actually take to raise an $11 billion fund?

Prior experience and relationships: the business started in 1988, so twenty years of good results built a reservoir of goodwill. The strategy suits crises, and they had managed through 1991 and 01-02 exceptionally well. They could point to flaws in the environment and the market failing as a disciplinarian, plus great respect for Bruce Karsh.

Answered around 15:23
What makes a business partnership last for decades?

Shared values and complementary skills. If one person is super aggressive and the other a chicken, or one is ethical and the other cuts corners, the partnership fails. Complementary skills make partners additive and synergistic rather than replaceable. The third element is being appreciative and thanking your lucky stars for a partner who does what you won't.

Answered around 22:17
How should a young person figure out what to do?

Find something that plays to your strengths, avoids your weaknesses, and makes you happy. That sounds obvious, but it means you can't let your friends, society, or your parents decide for you; you have to think it out yourself. It's difficult because it's hard to know yourself, and in twenty years you'll be a different person.

Answered around 30:07
What was Charlie Munger's greatest contribution to Warren Buffett?

Warren used to buy cats and dogs because they were cheap, what's called cigar butt investing: picking a used cigar out of the gutter for the three puffs left in it. Charlie's contribution, which most people credit as his greatest, was talking Warren out of that, convincing him to buy great companies at a good price.

Answered around 41:51
Key Quote

“The idea that you can give it a job and not tell it how to do it and it'll figure it out is really unique.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 2:48▶ listen at 2:48
Key Quote

“And if that's true, and if AI doesn't have hair on its neck, then maybe there's a role left for experienced investors with judgment.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 6:12▶ listen at 6:12
Key Quote

“Second level thinking basically says if you don't see anything different from everybody else, you can't possibly be superior.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 8:32▶ listen at 8:32
Key Quote

“The best time to invest is in a crisis. You can't raise money during the crisis because the news is so terrible.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 17:26▶ listen at 17:26
Key Quote

“A battle hero is not somebody who's unafraid. It's somebody who's afraid, but does it anyway.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 19:38▶ listen at 19:38
Key Quote

“And by the way, if you wait until you have nothing to be afraid about, probably the opportunity has passed.”

— Howard Marks, Howard Marks on AI and Human Judgment, at 20:14▶ listen at 20:14

Cite this

“Howard Marks on AI and Human Judgment.” https://dcs-spotify.megaphone.fm/HS7019324451.mp3. Transcript summary by WhipScribe, https://whipscribe.com/transcript-pages/howard-marks-on-ai-and-human-judgment.html.

Report this page
What is wrong with this page?