I would take that number with a huge grain of salt since the useless Google search ai box consumes Gemini tokens.

Intentional Gemini token consumption is likely very very small.

Any news on grannybench?

It's going to be all about the training set. Theres no moat in models trained on internet data.

There are tons of private datasets where a model will be fine tuned to provide useful domain specific responses.

And this is why I think companies will wise up and stop sending their information to the likes of Anthropic and OpenAI: they will not be able to maintain their edge if either one of them decides to defect and start using all of their uploaded user data to improve the model. So they have a massive incentive to break their own TOS.

That's the push factor. The pull is, say there's a giant repo like windows or Adobe. LLMs produce spaghetti code because the repo doesnt fit in the context and it wasn't in the training set. If Adobe fine tuned it the LLM would be much more useful.

This will be the future model for businesses. Buy a generalized frontier model and finetune it and make it available internally. This model offers a lot more business benefits than the current risk.

I'm running a max of 1M context locally. I think the next real move will be continuous learning/infinite context. That will be another game changer.

Infinite context requires Infinite memory.

It wasn't a theoretical thing, it was a practical thing. Just like when we talk about 'infinite tape' when looking at Turing machines. As opposed to 'clear boundaries to how much context you can have because that is baked into the model'.

The idea that noisy usage data is worth that much for training as to balance the fatal risk of the crime you describe, imo borders on a conspiracy theory.

I am under the impression that curated training pipelines are driving the gains, not utilizing random usage data in nefarious ways.

It's not just the maximum refresh rate, the variable refresh rate allows for power saving.

But can save more by dropping the refresh rate to 1hz and upping to 30, 60, 120 as needed.

Yup. gorgetbench, grannybench, and tebufranksalso bench are all completely unsaturated.

If they had charged people the economic price up front instead of intentionally bait and switching this wouldn't be controversial.

This is going to be the next oceangate and I fully support it.

Nitrogen specifically

I'll bite. What do you see as applications of for commodity cold gas thrusters? Excluding rockets.

Rapid adjustment of drone flighpaths with micro-thrusts? Getting interceptors out of tubes faster without explosions/temps/fumes? Anything where rocket propellant, heat is a problem?

Breaking, emergency braking in trucks? Kind of like reverse thrust in airplanes.

1. No rockets

Drones are extremely weight sensitive. Adding a heavy thruster would massively eat into the payload.

Trucks are also weight limited by the 80k rule. Brakes are far superior.

I accidentally went to a DSA meeting once. Was entertaining, they spent much of the time attacking eachother.

Putnam did followup work investigating the causes.

https://www.puttingourdifferencestowork.com/pdf/j.1467-9477....