• 2 Posts
  • 200 Comments
Joined 2 years ago
cake
Cake day: September 27th, 2024

help-circle
  • Shh, LLMs are quite good at stuff like this. Let them teach you. Use a command line harness like opencode or whatever tickles your fancy.

    Use a local model like Qwen 3.6 if you can, and start having fun. You might to have to learn how to use a LLM effectively though, to mitigate their disadvantages like hallucinating and their non-deterministic nature. Also, they want to do everything with curl, maybe find some alternatives to that.








  • Cloudflare, for instance, shared that Mythos Preview was particularly adept at exploit chain construction, which is basically spotting how several bugs can be used to create a series of attacks that do more damage than a single exploited flaw.

    But Anthropic also revealed that Mythos isn’t necessarily ready for primetime, which might also be part of why it’s actually keeping it to such a small and controlled base of users.

    Cloudflare also noted that other models found a lot of the same bugs as Mythos—an observation that has been made elsewhere. A security company called Aisle tested several small, open-source models and was able to find the same vulnerabilities that Anthropic highlighted when it announced Mythos—vulnerabilities that went unnoticed by humans for decades.

    My take is that Mythos is quite good, but is of course overhyped by Anthropic. Current models are quite capable now and experts can deploy them effectively. Good news is that a script-kiddie still will have a hart time to find and create workable exploits.


  • This is a question I keep asking myself about, well, the entire internet lately. “What if this all just doesn’t work this time?” What if AI companies can’t replace search? What if streaming video can’t replace Hollywood? What if short-form video apps can’t replace social media users? What if Silicon Valley can’t brute force their products — and ideology — onto the masses again? To take it even further, what if the dot com crash and the AI crash are actually part of the same 25 years epoch of technological stagnation? I’m not going to lie, I find this line of questioning exhilarating.

    What if OP is high on drugs? What if he just strings a list of hot takes together that are getting more and more nonsensical? What if the only reason he talks about technological stagnation is because he still hasn’t his flying car yet?? What if a meteor kills us all tomorrow??? I‘m not going to lie, I find this line of questioning mindnumbingly boring.