https://archive.org/details/dli.ernet.14756/page/3/mode/2up

The GPU build out will keep pumping tokens until the cost to replace is less that the cost to maintain. It is effectively sunk cost.

The story really didn't go the way I expected.

I feel like most people had an atlas or a paper map in their car most of the time. Though it is possible that I was more prepared than most.

What I haven't seen is any properly controlled study proving it false. Just a cluster of researchers in Tiawan with flawed study construction.

You make it a personal slight when you espouse, that your aesthetic preferences are universal and then farther malign people with different preferences by saying that they are not aesthecticly minded.

Here is probably one of the more clear examples. A model trained on video and also robotic simulation/recording (probably ensembled with control systems/mobility models) will likely be at the core of how robots make decisions and plan.

https://deepmind.google/blog/gemini-robotics-2-brings-whole-...

This is way outside my area of expertise though. I've only dabbled in more classic robotics and control systems, but these multi-modal sequence to sequence models are highly adaptive and can effectively transfer learning across very different domains.

We have only begun to extract the value of commoditized intelligence. Sure there are arguments on local models and pricing power.. but I think we will be compute constrained for the near future.

If I feel like the model and I explored a lot of options I won't want to keep context as it might be confusing.

I think the more you use it the better judge you are of whether you should purge, compress, or just keep the context before executing the plan.

The model would construct/hallucinate pre-conditions to satisfy some final output value that the model was predisposed to.

See how it fails or succeeds. Look at the supported features, try them out, think about how you might use them in your workflow.

Before you know it, you'll be proficient.

You have to learn how to self-teach.

So what you see is the market "stretching".. the bottom getting cheaper and the top end running away and getting more expensive. At some point the top end may be too valuable to even sell access to.

There I can listen to a 6 hours about the Olmec civilization.

I use this regularly for my personal financial research system. Even flagship models make mistakes. Though currently the issue is usually the model using a figure from and older report. Cross-check reduces that dramatically.

I think what people didn't realize was, just because the GPT-4.5 model didn't get better on the benchmarks, didn't mean the model wasn't different than the earlier models. It was being compared to thinking models that were being developed at the same time.

The GPT 4.5 model still has some of the most "human" like abilities in communication even though it isn't particularly good a problem solving. It hadn't under gone the same type of reinforcement training.

I still use GPT 4.5 sometimes, in creative exercises it can be surprisingly effective. The model is still available.

I'm hoping to see more work in the other direction with cyclic/looped transformers and other memory dense approaches.