NEW! ScottiesTech.Info presents: TekTalk

I watched an Asmondgold video as he watches another video claiming one AI released the code of another AI? I'm not savvy enough to get it but it seems significant. Is it? Apparently background code got to be seen by the public. Scottie, can you comment on this? (11:02)


This is the original video, posted five days ago, claiming with the summery line: "Anthropic accidentally leaked Claude Code's source code to the entire internet." (short: 7:21)

 
I watched an Asmondgold video as he watches another video claiming one AI released the code of another AI? I'm not savvy enough to get it but it seems significant. Is it? Apparently background code got to be seen by the public. Scottie, can you comment on this? (11:02)

I don't think that guy has any idea what he's talking about...

This is the original video, posted five days ago, claiming with the summery line: "Anthropic accidentally leaked Claude Code's source code to the entire internet." (short: 7:21)


That's more like it. Anthropic bought a company/tool that had a bug that accidentally resulted in the release of the source code for Claude. That's not AI releasing anything; that's just humans being dumb.

The significant thing is that now the world knows that Claude is not some ultra-amazing piece of coding, but a total mess of "hacks" to make a less-than-amazing LLM (the core of the AI) seem like it doesn't totally suck.

In other words, chances are most AI is actually just one giant hack - or at the very least, it's nowhere near as good as they claim. And that might explain the "hallucinations" and other quirks...
 
Episode 3:


Or listen to the podcast:

 
Back
Top Bottom