HackerTube

Andrej Karpathy — “We’re summoning ghosts, not building animals”

Dwarkesh Patel0 HN recommenders · 2025–2026
1.4M views · interview · dated · Uploaded 2025-10-17

Dwarkesh Patel interviews Andrej Karpathy on why reinforcement learning is flawed, why AGI may blend into steady GDP growth, why self-driving took so long, and the future of education. The conversation runs about two and a half hours.

Mostly cited as a source for Karpathy's views on agents, AGI timelines and LLMs, rather than recommended outright.

AI safety, AGI & impactMachine learning & deep learningTech pioneers & interviews

Watch on YouTube

5 comments from Hacker News

@moozilla2025 · in “Andrej Karpathy – It will take a decade to work through the issues with agents”

You could go with the title from the associated YouTube video ( [VIDEO] )? Andrej Karpathy — “We’re summoning ghosts, not building animals”

CitesView on HN
@jimbo8082025 · in “Tesla is heading into multi-billion-dollar iceberg of its own making”

This is kinda a hilarious coincidence, because I just watched a podcast with Andrej Karpathy, the guy who was running their autopilot program back when the sage "fully autonomous in one year" predictions kicked of their 15+ year long run (and counting). These days he's moved on to predicting 10 years for AGI and is, I shit you not, citing his 15 year track record of making accurate predictions (timestamped link below if you want have a laugh). [VIDEO]

CriticalView on HN
@bahmboo2025 · in “Karpathy on DeepSeek-OCR paper: Are pixels better inputs to LLMs than text?”

Not criticizing per se but I just watched this recent (and great!) interview where he extols how special written language is. That was my take away at least. Still trying to wrap my head around this vision encoder approach. He’s way smarter than me! [VIDEO]

CitesView on HN
@meken2025 · in “Backpropagation is a leaky abstraction (2016)”

He gives an update in the Dwarkesh interview: [VIDEO] Not too many details on timeline - just that he's working on it.

CitesView on HN
@graeber_289272026 · in “Ask HN: Why there are no actual studies that show AI is more productive?”

Electricity analogy is fairplay, but ChatGPT had something like 110% global adoption 5 minutes after its release. The infrastructure and the electrical appliances had to catch up, but the Internet is all built out already. So I think it's fair to be looking at results a few years in. Andrey Karpathy famously mentioned in an interview with Dwarkesh Patel [0], that the computer doesn't show up on GDP numbers, there's no noticeable jump or change in slope. Even if Excel is so damn fast, people are likely not drawing its full potential, and institutions are likely actively resisting change anyway. My take is that the general population hasn't found the productive levers yet, they're at the stage where they're happy to drag down and auto generate the date list in Excel, but don't know to adjust diagrams or read function docs, not to even mention VBS scripting. And the enthusiast (dev) community I'd say is starting adoption with internal tools, and shot-in-the-dark apps, but big successes need time to mature in all the other ways (design, reliability, user feedback, marketing...), which comes back to what you said also, that needs time. Product Market Fit isn't happening automatically by chance or good prompting, I would like to think. [0] [VIDEO]

CitesView on HN