Frontier AI Models Fail at Automated Research | Qasim Wani
MTS
November 6, 2026
Extrapolate co-founder and former xAI researcher Qasim Wani argues recursive self-improvement is fundamentally a data bottleneck rather than a model training problem.
Qasim Wani
For GDPX, we observed that the average accuracy or the mean reward for all the frontier models was not more than 24%, which means that there's still a lot of room to grow for these models to improve on. Part of the reason why these models are not good at it, and we sort of talk about it in our blog post within the GDPX report, one of the big problems with models like Astra or Carbon and Opus is that they don't know what experiments to prioritize.
SPEAKER_2
We're live with Qasim Wani, who's the founder of Extrapolate, which builds RL environments or co-founder of Extrapolate, which builds RL environments where AI agents learn to improve themselves by doing frontier research. And it's also contributed to GDPX, which is Seminalysis' new eval, which seems really cool and interesting. So, Qasim, welcome to MTS.
Qasim Wani
Thanks for having me.
SPEAKER_2
So, where should we start? We were just talking, before we went on, about the new meter report that says that, basically...
SPEAKER_3
Since October of last year.
SPEAKER_2
Yeah, since October of last year, the rate at which the Anthropic Epoch Capabilities Index benchmark score has been increasing is about double what it was previously. And they have other evidence that's non-public in addition, and that all this combined leads them to believe that Anthropics models are providing, like, substantial R&D uplift already.
So, do you have any takes on this? Like, if this line is true, like, does this mean we're entering RSI?
SPEAKER_3
What do you think it is?
Qasim Wani
It's one of those, like, if this, then that kind of situations. I think what's interesting to zoom out a little bit is to define RSI more concretely. There's something that we're talking about right before the show is different labs have a different perspective on RSI, and it's really interesting how a close friend of mine, he works at Google, and they claim that they're also working on RSI, but his team specifically does not work on model training. So it's like, what even is RSI at this point? What is RSI? It's the thing that's in vogue.
The way that I would say, I think about RSI is, can an agent invent the next version of itself?
You know, Anthropic in particular, their interpretation is from at least what I've heard. So internally, things could be a little different. But from what I've heard, their whole interpretation of RSI is, how good are models at doing pre-training research? And can agents automate all of pre-training? And that's sort of what Karpathy allegedly was hired to do at Anth was to automate a lot of pre-training research, which is absolutely one of the most critical things because everything is upstream of pre-training. But there's a lot more to self-improvement beyond just model training. And I think that's where we sort of get lost in the meta, which is why we created the RSI index with Semi-Analysis for GDPX.
And the biggest reason why we worked with Semi-Analysis on the RSI capability is because if you look at pretty much every single benchmark out there today, when it comes to AI research and this whole realm of auto research, RSI, a lot of them like Postrain Bench, Deep RSI, and the VALS RSI index, they're all hyper-focused on, can agents do some groundbreaking research? And while that is an important and a critical capability for models to be really good at, as Frontier Lab, you don't just tell Claude, hey, go build the next version of Claude. It's a lot more sophisticated and a lot more iterative by design. As researchers, we spend the majority of our time looking at the data. And so it's very iterative by design.
Researchers don't really have this idea of, I'm going to invent the next transformer. They're very iterative by philosophy. And what we see with a lot of these benchmarks, like let's say Post-Train Bench in particular, the benchmark has seven tasks. And one of the tasks is, can you take one of the open-source Quen models and post-train it to have the best performance on AME or GSM8K? The problem with those types of tasks is, that's not what you do inside the labs. Inside the labs, you're hill climbing on 20 benchmarks at once and not a single benchmark. And that was sort of like our version of RSI, is like, let's just do what we did at the lab. So I guess a little bit of background about me, I started the RSI efforts at XAI, working with Gudong Zhang, who was one of the founders of the company. And our philosophy when we started working on this was, to be able to have Grok invent the next version of Grok, it needs to be good at a multitude of things. It needs to be really good at model training, but it needs to be equally good at identifying what data to train on. It needs to be really good at creating the environments, doing harness design, eval design, infra. And those are like the primitives required to scale a lab. And so, that's where I think a lot of the meta gets lost these days, where people are like, oh, RSI is just kind of model train something, but it's actually much broader than that.
Read the other 16 minutes of this episode here.
$15 gets you this episode and 99 more. Credits never expire. After checkout, one link brings you back here to read it.
Not what you expected? Full refund within 14 days.
Thousands of transcripts fetched by people building searchable podcast archives
Fetch transcripts with code
The same credits work from a script or an AI agent. The demo key returns a sample episode in full, no card needed:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000800187186