woggy

I think I'd prefer NYT in Hackernews form

OK, so is this how LLMs will be able to learn?

Just had a thought: two local models, one with static weights trained mostly in the space of manipulating neural network weights, manipulating the weights of a dynamic LLM to enable learning.

Well they are much harder to read than tests.

I was going through an experiment trying to spec out a simple caching system in Lean. I was getting Gemini to help me run through doing this, step by step. At one point it writes out this statement with a proof to show something, the specifics of which I forget, but it was some simple property of the cache. I stare at it for a while thinking to myself this can't be right, it was some sort of plain tautological statement that didn't really say anything, like saying 1 = 1.

I give it to Opus 5.5 and it basically confirmed my gut feeling that it was useless.

I'd like to rely more on formal verification. I like the idea of starting out with specs in a formal setting and designing software that way. The problem is it's hard to do right, and my point is even with AI, it's hard to do right.

I found that even with LLMs making sure the formal specification covers what you actually want is challenging.

How do you actually know what the LLM proved is what you want?

Is this Zucks personal AI bot?

I would replace the AI avatars with Daedalus, Icarus, Helios and Morpheus avatars from Deus Ex

Have you guys done any talks or written any articles on how you are developing this software?

Scaling up to this kind of complexity even with AI is still a challenge, at least for me.

I’m curious whether people can use Lean primarily as a software specification language, without necessarily intending to prove everything.

Can we use it to specify module meanings and laws, to make the design precise and checkable, and would allow us to implement property based tests for those laws in our implementation. Lean becomes a tool for more precise thinking about the design.

They should only make these announcements just before releasing. Or at least put a date to it.

I prefer Pro Max Ultra XXL

I can't remember exactly what I did but I think passwords for windows 2000 were stored in something called a SAM file. I managed to get a copy of it and brute force it on my home machine. Password was the schools phone number. Got access to the whole network, network drives, teachers files. I was paranoid so I didn't poke around too much :)

They can keep the buttons and the normal interiors instead of making them look like something out of Star Trek.

Is it that hard? Take a normal car, take engine out, put electric motor in. Don't fuck around with the inside.

Plans for Windows version?

And how lucky Israel allowed and facilitated the transfer of hundreds of millions of dollars to them.

US should mind its own business

"When 1 + 1 + 1 Does Not Equal 3

Consider the number 37. On its own, it is merely data. Say that it represents millions of dollars of investment exposure: that adds context."

Who can even read this? People, please, put some effort into what you publish.

I don't understand this sentence, can you try again please? Are you saying Laya was built on research done by the Jev team?

Anyone using tools like ndepend or others to help guide agents in refactors?

Personally I have a some tools that build dependency graphs (C# and Python) and store the results in a local database. Agents seem quite good at poking at this and coming up with refactor ideas. Graph analysis tools are useful here, simple application will detect cyclical dependencies, but I encourage the agents to use more complex tools like clustering to poke at the data.

Can this be used in conjunction with a text-generating LLM for better quality code generation?

Sunday Robotics has a beta program planned for end of year

I don't really know much about this stuff, but are systems available that can detect this kind of behavior and react accordingly?

Yes, this is OpenAI hacking other entities ... clearly this is on OpenAI

Interesting, never heard of this. But it looks so, so wrong.

MCPs are useful in an engineering context to allow for API access in applications where the API needs to run in-process. e.g. in Autocad funnel a script in, compile it and run it within the Autocad environment. The MCP does little more than just take scripts in and return results out, and can handle multiple active target processes.

Bobcoins. Hehe.

I guess Ontario needs to be renamed to America now ...

I think this is a lot better from how I remember the previous site; it doesn't immediately come off as LLM generated.

I'm not sure, when I try to picture something I just get the concept of it in my head. I thought people were just really good at taking that and working out how to draw it out. Makes a lot of sense to me now that some people can be so good at drawing.

Why maximum notice? Why not minimum notice? Elon doesn't deserve more than that.