mashlol

What if it depends on licensed code that you cannot legally distribute?

For me, Firefox on desktop (Windows) and mobile (Android) both give the SSL warning. Although mine appears a bit different than the parent comment:

    An error occurred during a connection to haxxorwpm.0s.is. SSL received a record that exceeded the maximum permissible length.

    Error code: SSL_ERROR_RX_RECORD_TOO_LONG

Do you have an example of someone doing that? I suppose the argument would be we wouldn't know - but personally I don't buy it. It's possible to not be "famous" as a billionaire, but it's not possible for people in your life not to know.

Even if only 5% of his NW is in cash, that'd be $50billion dollars in actual liquid cash. Even if it's 1% or less, he almost certainly has over $1bn in cash or practically liquid cash. That's already a dragon's hoard.

Didn't AOC's comment specifically say a billion? Why would it imply someone increasing from $200k to $10m didn't earn it?

I usually make the llms do that part for me. Instead of asking the llm to refactor, ask it to write the codemod script that'll refactor, have it test that script, and even have it run it on its own. It's definitely faster and less error prone that way for me.

Has rate of progress increased? How does one measure that? Genuinely curious - would be very interesting to map out the "effectiveness" of each AI model vs how long it took to train/release.

From my perspective, the model gains are mostly incremental now and a lot of the gains are just from things like improving the agent harnesses. I could be wrong though.

If you try to just instruct them they might get it wrong. If the surrounding software forces them to do it, it'll always work. E.g. it can check for merge conflict markers like <<<<<<< and re-invoke codex/claude to merge again if the previous resolution failed for whatever reason (e.g. AI hallucinated, threw up, whatever).

Also you'll need a wrapper to actually detect when merge conflicts will occur and when rebasing is necessary.

Generally, the less you rely on the AI the better. Make the AI write the code, sure, but don't make the AI be the process.

Why Mac only?

Also - one issue I've seen with other tools doing worktree stuff is they don't deal with merge conflicts automatically. IMO the agents should just automatically resolve conflicts & rebase on their own, is that a thing here?

I'm not an expert, but as I understand it there are existing solvers for poker/holdem? Perhaps one of the players could be a traditional solver to see how the LLMs fare against those?

AI almost always reduces the time from "I need to implement this feature" to "there is some code that implements this feature".

However in my experience, the issue with AI is the potential hidden cost down the road. We either have to:

1. Code review the AI generated code line by line to ensure it's exactly what you'd have produced yourself when it is generated or

2. Pay an unknown amount of tech tebt down the road when it inevitably wasn't what you'd have done yourself and it isn't extensible, scalable, well written code.

> which is what every other major studio would have done in its place

Afaik CDPR doesn't make many games. If one flops, that might be the end of them. I don't see abandoning a game as a valid option for them from a financial perspective. Makes much more sense to fix the issues and sell more.

Would you say the same if someone played 1000 hours of a sport?

Definitely seems like it could be useful, but I'd be worried with giving AI write access to emails.

Is there a good audit trail of exactly what actions it takes at each step? I'd personally be worried about leaking proprietary or otherwise private information this way, or having it hallucinate information when it sends out emails potentially causing catastrophic issues.