Tem kx

[Tem kx @Temkox]

@wheresryan22 @kzmucx @SpaceX commenting for the algo to bring me backk!

[Tem kx @Temkox]

@_overment My GOAT! ty so much. Will def be trying it out this week

[Tem kx @Temkox]

@CesarDemi81 @mrownz123 @YGandelsman brother respondiste una repuesta de un comentario de twitter que nadie te etiqueto, jajaj. Vos te metiste solito

[Tem kx @Temkox]

@grok @sevensix43 @CesarDemi81 @mrownz123 @YGandelsman @CesarDemi81 se ve que ni siquiera es verdad que "GLM defended them" cuando no paso. Lo que era "widely know" era que no se pudo usar fue para hacer el analisis forenstics, no para defenderlos

[Tem kx @Temkox]

@fd_ripatel @YGandelsman @SCherian5734 And even if you would claim that the real risk is not intra-country hacking but inter-country (china vs us). If the real adversary is China. That would be even more incentive for the US to ban Chinese models (not saying I agree or believe it should be done)

[Tem kx @Temkox]

@fd_ripatel @YGandelsman @SCherian5734 But then if the models (that are so good at detecting and preventing cybersecurity usage as we are saying) are not allowing hackers to exploit it, then which other models are left for hacker to exploit it?

[Tem kx @Temkox]

@grok @sevensix43 @CesarDemi81 @mrownz123 @YGandelsman @grok did they use or tried to use an OpenAI/Anthropic model to cut the attacker access?

[Tem kx @Temkox]

@CesarDemi81 @mrownz123 @YGandelsman entonces porque le dirias que es "widely known" cuando efectivamente esa persona no lo sabe. Cuando tu hijo te pregunta porque sale el sol, no le decis "its widely known bro". Media rara forma de comunicarte... con la gente...

[Tem kx @Temkox]

@oofdere @fd_ripatel @YGandelsman @SCherian5734 we are not talking about ai here, we are talking about the specific guardrails in cybersecurity and its effect on hacking/safe usage. the concept-object here is 'the specific set of skills/services of ai used for cybersecurity', not 'ai models'

[Tem kx @Temkox]

@oofdere @fd_ripatel @YGandelsman @SCherian5734 ¿Or would you say that using a knife to stab someone isn't a net negative?

[Tem kx @Temkox]

@oofdere @fd_ripatel @YGandelsman @SCherian5734 so what you are saying is that is, an usage of ai for hacking purposing isnt a net negative?

[Tem kx @Temkox]

@CesarDemi81 @mrownz123 @YGandelsman brother I asked because I didnt know..... Not because I was making an ironic/jab comment

[Tem kx @Temkox]

@fd_ripatel @YGandelsman @SCherian5734 You have to look at it from a lens of liability more than a free creation of a tool/tech like a calculator, the press, etc. The AI Labs are giving you a service, and that service cant be used for harm. Even tho maybe 99% of its usage is virtuous/safe

[Tem kx @Temkox]

@fd_ripatel @YGandelsman @SCherian5734 You have to look at it not in terms of normal usage but systemic threat. The 90% of usage of guns are from policemen. That doesnt prevent the 10% of killers from killing plenty of people. There are a lot analogous examples, of companies that could/may harm people with its tools.

[Tem kx @Temkox]

@oofdere @fd_ripatel @YGandelsman @SCherian5734 I am referring to the analogous "virtuous" usage of a gun, in terms of defending yourself from being killed for example. Not saying I believe that is a virtuous action/behaviours

[Tem kx @Temkox]

@mrownz123 @YGandelsman this is a good point, is this confirmed?

[Tem kx @Temkox]

@fd_ripatel @YGandelsman @SCherian5734 "Great, now that everyone has a gun, I can use it for my safe/virtuous usage."

[Tem kx @Temkox]

@ClaudeCodeLog great video, love this style

[Tem kx @Temkox]

@shreyacasmalert fire! thanks!!