Not if your use case needs speed. For one of my products I can't use an LLM that has a p99 of >700ms for TTFT.

If it could output 1k tokens per second but needed 4 seconds to produce the first batch of 4k, would that not be viable?

They need to catch up to OpenAI, so it makes sense to skip a few numbers.

Which doesn't track. Public transport in Japan is peaceful. I feel more relaxed on a Train in Japan than a car or Train in Canada.

The PRC has made threats towards Japan. It makes sense to consider them in the context of Japan defending itself.

Very cool! I created something similar (with the same name), but in 2 tools. One to share skills (skillsync) and another to share context (contextsync).

I like to create a design with one model, sync it to another and iterate on it, then potentially go to another to apply/implement.

Curious why you didn’t bundle them?

I did this, kind of, with my companies internal tools. LLM<->LLM links.

That's actually pretty neat.

>but if you're doing fundamental research it's 99% stuff you are building yourself with your own hands.

You can do LLM->3D Printed models now. The drone can fly in and pick them up and bring them to the location you want. They can assemble structures. All automated, all LLM driven.

Things are changing. What was true, no longer is.

With much higher energy costs + consumption of consumables like battery packs than just driving a truck loaded full of building materials.

I left AWS recently, but I don't think the entire place is a mess. It really depends on the team.

On my team we were expected to work 90+ hours a week, and were told we were slackers if we were unwilling to work Sunday afternoons. But I've heard from others that some teams are a lot better.

Many Amazon teams have had high expectations for work. The stories I’ve heard have largely been around how AI is used, slipping CR standards, and what is funded.