wobbat

AI & Security. Hacker, Senior Offensive Security Consultant and sometimes human. Creating, breaking, and improving systems.

[wobbat @w0bbat]

If we just keep saying ‘Fable 5 is coming back today’ one of these days we gonna be right.

[wobbat @w0bbat]

@thdxr I have early access to Le Chaton Fat. Don’t worry open source models gonna be real good. Fable who?

[wobbat @w0bbat]

@argofowl

[wobbat @w0bbat]

@roohaa_n It is not bad, but still GPT 5.5 provides a lot better results for me.

[wobbat @w0bbat]

@bridgemindai But have you guys tried Le Chaton Fat?

[wobbat @w0bbat]

@matt503ea5sf9z5 Fable 5 would make a good Anthropic CEO

[wobbat @w0bbat]

@mitsuhiko I am not a windows user of pi, but I wanna assume the biggest pain point is going to be windows.

[wobbat @w0bbat]

@Otiscode @OpenAIDevs Click on the goblin sweater? 😉

[wobbat @w0bbat]

@MattSchrage I did not know that was a thing. I will have to check that out! I would love to have a chance of trying swe fast more seriously with code review!

[wobbat @w0bbat]

@cognition Would love to try!

[wobbat @w0bbat]

@cognition @cerebras Pls!

[wobbat @w0bbat]

@MilksandMatcha @cerebras I’m a senior security professional trying to push models as far as I can helping open source projects. I’m attempting this (just started) by building an open source AI based security code scanner. Would love to have a chance to get to try out windsurf for this!!

[wobbat @w0bbat]

@Dimillian Arch! BTW I use arch!

[wobbat @w0bbat]

@thsottiaux When will we get the app for Linux? 🤞

[wobbat @w0bbat]

@thdxr AI will, and already clearly does, convince many they have found zero click full RCEs everywhere, without being able to validate an overly helpful AI proving a non existent vulnerability. It’s going to be so much worse.

[wobbat @w0bbat]

@thdxr As someone who does security assessments and code reviews all day, I think AI will make security much better, and some people much more annoying. AI slop in security reporting is the worst. Peeps chasing five minutes of fame is a sad side effect of the security field.

[wobbat @w0bbat]

@adamdotdev Yes, I have not had anything as crazy as the 4,5 hours session you had this morning, but I am seeing quite a bit longer thinking times. The thinking seems reasonable, just a whole lot it. I am using zen and opencode as well though.