Episode summary
Is AI Slowing Down? Nathan Labenz Says We're Asking the Wrong Question
The a16z Show
· · Hosts: Erik Torenberg · Guests: Nathan Labenz
Summary
Nathan Labenz argues that the question is not whether AI is “slowing down” in any absolute sense, but which part of the system is improving fastest. He agrees that some people are becoming lazier with AI and that tools can reduce cognitive effort, but he rejects the idea that progress has flatlined or that GPT-5 is only a small step beyond GPT-4.
He says much of the apparent disappointment around GPT-5 came from launch problems, confusing model routing, and expectations that were set too high. On his reading, the real story is that improvement has shifted from pure scaling toward post-training, longer context windows, and extended reasoning. He points to gains in long-tail knowledge, math, and scientific problem-solving, including IMO-level performance and AI systems that are beginning to help discover new results in science.
Labenz also says the economic impact is already visible in customer service, document review, and software, even if it is uneven. He expects fewer engineers over time, substantial disruption in jobs, and more automation in multimodal areas like robotics and self-driving cars. His broader thesis is that AI should be understood as a general technology stack, not just chatbots, and that the next phase could look more like systems solving real-world engineering problems than answering prompts.
Key takeaways
- Nathan Labenz says the “AI slowdown” debate confuses user experience with capability progress, and he thinks GPT-5 is still a meaningful leap over GPT-4.
- Labenz argues that launch perception mattered: broken routing and high expectations made GPT-5 look weaker than it was.
- He says extended reasoning and long context are now as important as scaling, and they are producing real gains in math and science.
- Labenz expects job disruption to show up first in high-volume, repeatable work like customer service, document review, and some software tasks.
- He thinks AI is moving beyond language models into vision, biology, robotics, and self-driving systems, which makes the job impact much larger than chatbot use alone.
Notable quotes
“AI is not synonymous with language models”
“I think you start to have something that looks kind of like super intelligence”
How the episode opens
AI is not synonymous with language models. AI is being developed with pretty similar architectures for a wide range of different modalities and there's a lot more data there. Feedback is starting to come from reality.
Maybe we're running out of problems we've already solved. When we start to give the next generation of the model these power tools and they start to solve previously unsolved engineering problems, I think you start to have something that looks kind of like super intelligence. Nathan, I'm stoked to stoked to have you on the AC and Z podcast for the first time.
Obviously, I've been podcast partners for a long time with with you leading Cognitive Revolution. Welcome. It's great to be here.
Thank you. So, we we were talking about Cal Newport's podcast appearance on lost debates and and we we thought it was a good opportunity to just have this broad conversation and really entertain this sort of question of is AI slowing down? Um, so why don't you sort of steelman some of the arguments that you've heard on that side either from him or more broadly and then we could sort of have this broader conversation.
Yeah, I mean I think for one thing um it's really important to separate a couple different questions I think with respect to AI. One would be is it good for us you know right now even and is it going to be good for us in the big picture? And then I think that is a very distinct question from are the capabilities that we're seeing continuing to advance and you know at a pretty healthy clip.
Um so I actually found a lot of agreement with the uh Cal Newport podcast that you shared with me when it comes to some of the worries about the impact that AI might be having even already on people.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai slowdown
- gpt-5 debate
- scaling laws
- reasoning and automation
- ai agents and coding
Watch the episode
Watch “Is AI Slowing Down? Nathan Labenz Says We're Asking the Wrong Question” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from The a16z Show
-
How AI Could Bring Halo and Other Classic Games Back to Life -
Mike Volpi on Why AI Infrastructure Will Outlast the Hype Cycle -
Why HubSpot’s Co-Founder Is Betting Against the SaaSpocalypse -
Inside Salesforce’s Acquisition Strategy for the AI Era -
How AI Is Turning Figma’s Finance Team Into Builders -
The New AI Agents That Shop, Plan, and Gossip
The a16z Show, this episode and its artwork belong to Andreessen Horowitz. Octagon is not affiliated with the show; this page is an independent summary.
