hendler

Google gears is grinding

Cloudflare is certainly making a full court press to provide infra for the agentic web. My own ability to see whats coming also believes we need wallets, agent protocols, cloud services for agents, etc. So, maybe this is the way.

At the same time, something tells me what is coming isn't going to be as recognizable as these products build for.

A company's values become even more important with AI accelerating and expanding the mission of traditional Delaware Corporations.

Do you have advice on how to use AI to help teams stay true to their values?

Having not read your book yet, in my mind there's the obvious legal support AI can provide to help navigate complex situations, but maybe there's some other groundwork in the value creation and implementation itself?

...And Semantic Web.

Just ask, "I need to wash my car. If a carwash is 50 ft away should I walk or drive?"

There's some interesting, sad, but hopeful science fiction about where this is headed.

Ministry for the Future: https://www.goodreads.com/book/show/50998056-the-ministry-fo...

Excerpt here: https://orionmagazine.org/article/the-ministry-for-the-futur...

With the TPU deal with Google and their relationship with Amazon they will have access compute coming online.

I worked with 4.6 and found some improvements for better planning and sustained us, but agree some posters 4.7 is slower, overthinking.

What I expect is frontier models to get bigger and more expensive (especially fast mode like on Cerberus). And most of his get much smaller distillations for the more generous subscription tiers.

The question is not if the commenter is an expert, but if they are correct.

The claim that some models didn't take larger systems into account is also because an expert in the arctic wasn't an expert in oceans. And the expert in biodiversity isn't an expert in food supply chains. Expertise isn't the question. Instead it is - do all of us who are non experts (all of us) have enough expert data to have a systemic understanding of an accelerating trend?

same for me

Useful for embedded devices? Crashes, disk updates not important for ephemeral process?

Since this is open source, it's a white-hat tool, but it also democratizes script kiddos being able to do some serious damage.

It's happening everywhere, all at once. I guess who owns the most hardware wins the arms race?

Finland happiness study revealed the concept of "go first".

HTML was XML for the web. Nothing to it from a technical perspective.

Check out BaseTen for performant use of GPUs

Google TPUs, Amazon, a YC funded ASIC/FPGA company, a Chinese Co. all have custom hardware too that might scale well.

Is this because 405b doesn't fit on Groq? If they perform better, I would also have liked to have seen.

Seems like a nice abstraction.

Since I see DuckDB mentioned, folks wanting serverless may also be interested in LanceDB, written in Rust, with most features built out for Python.

https://lancedb.com/

https://github.com/lancedb/lancedb

Side note, I wrote a proof of concept of embeddings generator being handled inside PostgreSQL, independent of the index.

https://github.com/Hendler/flame

Thanks! Feel free to reach out.

Your answer is part of why I am concerned. Maybe new incentives or leverage can be created. Giving AI the power to create new social dynamics seems both too dangerous and necessary.

I care a lot about the environment, and worked in climate tech for 7 years. I also am very hopeful for AI's contributions.

When we say that AI will help save the environment, I can see short term gains in things like nuclear energy expansion and improvement, better educational opportunities, robotic labor and other efficiency automations, etc. On ecosystems and biodiversity, I don't see a clear path.

AI can improve our understanding, but a systems collapse requires very complex bio-chemical systems that are all interdependent. Maybe we all live in bubbles of AI controlled environments post Holocene, but we rely on larger earth systems that will be difficult to rebuild post collapse.

Does anyone else worry that AI will not be rapid or sufficiently impactful to deal with ecosystem collapse?

I also wrote wrote an LLM to Prolog interpreter for a hackathon called "Logical". With a few hours effort I'm sure it could be improved.

https://github.com/Hendler/logical

I think while LLMs may approach completeness here, it's good to have an interpretable system to audit/verify and reproduce results.