logo
episode-header-image
Jun 24
2h 48m

#219 – Toby Ord on graphs AI companies w...

Rob, Luisa, and the 80,000 Hours team
About this episode

The era of making AI smarter just by making it bigger is ending. But that doesn’t mean progress is slowing down — far from it. AI models continue to get much more powerful, just using very different methods, and those underlying technical changes force a big rethink of what coming years will look like.

Toby Ord — Oxford philosopher and bestselling author of The Precipice — has been tracking these shifts and mapping out the implications both for governments and our lives.

Links to learn more, video, highlights, and full transcript: https://80k.info/to25

As he explains, until recently anyone can access the best AI in the world “for less than the price of a can of Coke.” But unfortunately, that’s over.

What changed? AI companies first made models smarter by throwing a million times as much computing power at them during training, to make them better at predicting the next word. But with high quality data drying up, that approach petered out in 2024.

So they pivoted to something radically different: instead of training smarter models, they’re giving existing models dramatically more time to think — leading to the rise in “reasoning models” that are at the frontier today.

The results are impressive but this extra computing time comes at a cost: OpenAI’s o3 reasoning model achieved stunning results on a famous AI test by writing an Encyclopedia Britannica’s worth of reasoning to solve individual problems at a cost of over $1,000 per question.

This isn’t just technical trivia: if this improvement method sticks, it will change much about how the AI revolution plays out, starting with the fact that we can expect the rich and powerful to get access to the best AI models well before the rest of us.

Toby and host Rob discuss the implications of all that, plus the return of reinforcement learning (and resulting increase in deception), and Toby's commitment to clarifying the misleading graphs coming out of AI companies — to separate the snake oil and fads from the reality of what's likely a "transformative moment in human history."

Recorded on May 23, 2025.

Chapters:

  • Cold open (00:00:00)
  • Toby Ord is back — for a 4th time! (00:01:20)
  • Everything has changed (and changed again) since 2020 (00:01:37)
  • Is x-risk up or down? (00:07:47)
  • The new scaling era: compute at inference (00:09:12)
  • Inference scaling means less concentration (00:31:21)
  • Will rich people get access to AGI first? Will the rest of us even know? (00:35:11)
  • The new regime makes 'compute governance' harder (00:41:08)
  • How 'IDA' might let AI blast past human level — or not (00:50:14)
  • Reinforcement learning brings back 'reward hacking' agents (01:04:56)
  • Will we get warning shots? Will they even help? (01:14:41)
  • The scaling paradox (01:22:09)
  • Misleading charts from AI companies (01:30:55)
  • Policy debates should dream much bigger (01:43:04)
  • Scientific moratoriums have worked before (01:56:04)
  • Might AI 'go rogue' early on? (02:13:16)
  • Lamps are regulated much more than AI (02:20:55)
  • Companies made a strategic error shooting down SB 1047 (02:29:57)
  • Companies should build in emergency brakes for their AI (02:35:49)
  • Toby's bottom lines (02:44:32)


Tell us what you thought! https://forms.gle/enUSk8HXiCrqSA9J8

Video editing: Simon Monsour
Audio engineering: Ben Cordell, Milo McGuire, Simon Monsour, and Dominic Armstrong
Music: Ben Cordell
Camera operator: Jeremy Chevillotte
Transcriptions and web: Katy Moore

Up next
Jul 8
#220 – Ryan Greenblatt on the 4 most likely ways for AI to take over, and the case for and against AGI in <8 years
Ryan Greenblatt — lead author on the explosive paper “Alignment faking in large language models” and chief scientist at Redwood Research — thinks there’s a 25% chance that within four years, AI will be able to do everything needed to run an AI company, from writing code to design ... Show More
2h 50m
Jun 12
#218 – Hugh White on why Trump is abandoning US hegemony – and that’s probably good
For decades, US allies have slept soundly under the protection of America’s overwhelming military might. Donald Trump — with his threats to ditch NATO, seize Greenland, and abandon Taiwan — seems hell-bent on shattering that comfort.But according to Hugh White — one of the world' ... Show More
2h 48m
Jun 2
#217 – Beth Barnes on the most important graph in AI right now — and the 7-month rule that governs its progress
AI models today have a 50% chance of successfully completing a task that would take an expert human one hour. Seven months ago, that number was roughly 30 minutes — and seven months before that, 15 minutes. (See graph.)These are substantial, multi-step tasks requiring sustained f ... Show More
3h 47m
Recommended Episodes
Jan 2025
With OpenAI seeking profits, activist seeks payback to the public
A battle is brewing over the restructuring of OpenAI, the creator of pioneering artificial intelligence chatbot ChatGPT. It was founded as a nonprofit in 2015 with the goal of developing AI to benefit humanity, not investors. But advanced AI requires massive processing power, whi ... Show More
15m 21s
Oct 2024
OpenAI’s Path to Become a For-Profit Company Is Complicated
OpenAI plans to convert from a non-profit to a for-profit organization, a complex move that is rarely done. WSJ reporter Theo Francis joins host Zoe Thomas to discuss the hurdles that OpenAI will face and the possible reasons for the change. Plus, a controversial bill to regulate ... Show More
13m 16s
Oct 2024
20VC: Why Founder Mode is Dangerous & Could Encourage Bad Behaviour | Why Fundraising is a Waste of Time & OKRs are BS | Why Angel Investing is Bad for Founders to Do and the VC Model is on it's Last
Zach Perret is the CEO and Co-Founder of Plaid, a technology platform reshaping financial services. To date, Zach has raised over $734M for Plaid from the likes of NEA, Spark, GV, Coatue and a16z, to name a few. Today, thousands of companies including the largest fintechs, severa ... Show More
50m 53s
Mar 2025
Nonprofits navigate Trump’s drastic funding cuts, with The Chronicle of Philanthropy CEO Stacy Palmer
President Trump’s dramatic cuts to U.S. government grants are destabilizing every corner of the non-profit sector, leaving organizations scrambling to adapt. Stacy Palmer, CEO of the Chronicle of Philanthropy, explores the executive order’s impact on both the public and private s ... Show More
27m 37s
Apr 20
Purpose Isn’t Found, It’s Built with Aaron Hurst
In episode 230 of The Business Development Podcast, Kelly Kennedy is joined by Aaron Hurst, bestselling author of The Purpose Economy and a pioneer in the field of meaningful work. Together, they explore the idea that purpose isn’t something we find, but something we intentionall ... Show More
1h 5m
Dec 2024
Fixing Education in America: What's Stopping Us?
Over half of Americans live in childcare deserts, while 90% of brain development happens before the age of five. All the while, education and childcare remain among the most resistant sectors to technological change. Billions of dollars have been spent, but outcomes continue to l ... Show More
38m 54s
Dec 2024
Sam Altman's Equity Dilemma
In this episode of the AI Chat podcast, host Jaeden Schaefer discusses the controversies surrounding OpenAI, particularly focusing on Sam Altman's claims about his equity in the company. The conversation delves into Altman's testimony before Congress, the significant valuation gr ... Show More
14m 4s
Nov 2024
ChatGPT’s First Victim + The Department of Government Efficiency (DOGE)
Scott and Ed open the show by discussing Spotify and Disney’s earnings, a gambling company’s strong third quarter results, and Elliot Management’s activist investment in Honeywell. Then Scott breaks down how Chegg allowed ChatGPT to take its business to the woodshed and why he th ... Show More
54m 57s
Jul 2024
Pump and Dump Schemes are Now Totally Legal
Send us a textA Texas District Judge Andrew S. Hanen has dismissed all charges against seven social-media influencers the SEC and Justice Department had accused of perpetrating a “stock manipulation scheme” on Twitter and Discord, ruling that the prosecution failed to state an of ... Show More
20m 27s