Jacob Morrison

PhD student @uwnlp @uwcse, research @allen_ai @ai2_allennlp

[Jacob Morrison @jacobcares]

I'm so excited that we're finally releasing Tülu 3, our new post-training recipe! We're releasing models built on top of Llama 3.1 base (OLMo coming soon!), all of our datasets, a (73 page!) paper, new evaluations, and all of our code.

[Ai2 @allen_ai]

Meet Tülu 3 -- a set of state-of-the-art instruct models with fully open data, eval code, and training algorithms.

We invented new methods for fine-tuning language models with RL and built upon best practices in the community to scale synthetic instruction and preference data.

Demo, GitHub, technical report, and models below 👇

[Jacob Morrison @jacobcares]

@benmaritz true, and we haven't even gotten a real taste of the models co-built for bespoke inference hardware

[Texas Observer @TexasObserver]

From Daniel Dawer and Ruth Kravetz: Nearly 2,000 juniors at 16 high schools in Houston ISD were pulled from Algebra II and placed in remedial math. At some schools, about 70% of juniors were similarly demoted.

It wasn't due to poor grades ... https://t.co/zuyvh0CZwg

[Jacob Morrison @jacobcares]

@LeonDerczynski did you move back to seattle :O

[Jacob Morrison @jacobcares]

@ArmandDoma @tszzl when you have really really big ones

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ericzelikman @YuchenHe07 @parth_29 maybe 4 months

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ericzelikman @YuchenHe07 @parth_29 good, currently poisoning my mind having sol do all my coding for me (haven't opened an ide in 6 months)

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ericzelikman @YuchenHe07 @parth_29 we love the mind poisoner don't we folks

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ericzelikman @YuchenHe07 @parth_29 tbh would I have been better if somebody gave me a coding agent at my first few jobs back in the day

no

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ericzelikman @YuchenHe07 @parth_29 lmaoo

[Jacob Morrison @jacobcares]

@BenDLaufer welcome!

[Andy Masley @AndyMasley]

The reason data centers evaporate so much water is to maintain the cloud

[Taylor Lorenz @TaylorLorenz]

Hank Green self cancelling for using AI literally just for research and then using his platform to push dopamine nonsense pseudoscience AI addiction slop... lord help us!!!!!!

[Jacob Morrison @jacobcares]

@soldni noooo! glad mittens and/or mj are ok

[Adam Hunt @RealAdamHunt]

Recently I've flipped from being bullish to being bearish about AI.

I think I'm updating my bearishness to be more solidly bearish. Early thoughts (which I hope to be disproven in the next year or so, I would prefer progress) and my reasoning:

The whole 'it turns out if you keep training and scaling the models more they develop broad new capabilities in lots of domains' thesis is wrong (sorry Demis). The recent batch of models haven't got more general, they've got less general. This is most obvious in the fact that their language outputs have got much worse in comparison to e.g. o3. If they were gaining generalist capacities we would expect them to be describing their work in ever more graceful and comprehensive prose!

The image that was being shared as the AGI thesis (November 2025, Tomas Pueyo) was the spiky bubble that has a current spike or two out past human capabilities (e.g. on coding or math) but below human on other capabilities on the other spikes - the future prediction was that as the models scale/advance, every spike would grow bit by bit until the whole center encompasses the human capabilities, with super-superhuman on some spikes. I think it seems like what's actually happened in the last few models has been that the coding/math spike has grown, but leaving behind or even at the cost of the other spikes. The models are no better at some simple logic, language (and sometimes worse!).

This makes sense from a simple RL perspective; you can't RL something endlessly on one domain of tasks and expect it to improve on the other tasks. The fact that early LLMs did seem to improve generally was a byproduct of the written language corpus covering everything - that corpus is general, so training it on that gave the appearance of something generally intelligent and becoming more generally intelligent as it got better at replicating that corpus. But the actual logic and underlying ground truths behind the language aren't captured efficiently enough and weren't effectively RLd in - they top out at some point (I guess this happened around the time that there was the 'has scaling hit a wall' discussion in late 2024). Chain of thought was then a genuine breakthrough, along with web search, which plugged into that general LLM global-corpus intelligence to lead to post 2024 gains.

The AI companies have since worked out that coding works (and pays) really well (basically this is because the entire job is nearly perfectly recorded and exists as training data, and you can set up clear benchmarks and rewards). The recent models (and benchmarks) have been maxxing that and we've seen degradation on normal English use for that reason. This could still be transformative, leading to extremely powerful (and potentially dangerous, particularly in cyber security) models but it's not a pathway to AGI.

I'm probably at about 40% confidence about this. It fits my current observations of AI progress and has a basic explanatory model. It doesn't account for potential breakthroughs, which is a major reason for discounting.
To make some predictions, I guess if I'm right this will become broadly apparent and more widely acknowledged in the next year or two, as we see how the spikiness of models that keep getting released develops.
Maybe there will be efforts to concentrate on specific spikes e.g. health or law which require going back to earlier models and RLing on a different data set/with different rewards/benchmarks. Maybe those separate models can be linked together to give a more apparently general model. How capital intensive that is/the potential profitability will be a defining question. But I just don't see general abilities emerging atm, and I don't think we will any time soon. Good news - a whole industry of tackling important specific problems/sectors can open up!

[Jacob Morrison @jacobcares]

@saurabh_shah2 @baneepbanana the tech to existential dread to marathon pipeline stays undefeated

jk nice job

[Jacob Morrison @jacobcares]

@finbarrtimbers @codeviking

[Jacob Morrison @jacobcares]

@saurabh_shah2 @ethanlshen did they actually go to the conference

[Jacob Morrison @jacobcares]

@saurabh_shah2 8 days is the scientifically identified amount of time to feel refreshed, so a long weekend probably won't be enough (we've made that mistake before; it's ok to disconnect longer than you think)