0xRok0n

[0xRok0n @0xRok0n]

@FonsecaKaufmann @MiguelNicolelis @mehdirhasan I mean sure, send me that stuff. I’d like to have a look.

[0xRok0n @0xRok0n]

@FonsecaKaufmann @MiguelNicolelis @mehdirhasan The “incident” isn’t even real… If you read the original article plus watch the podcasts with some of this involved you’ll realize the incident was actually fabricated.

[0xRok0n @0xRok0n]

@ModerateNut @realBigBrainAI Bold of you to assume that quantum computers will simply run today’s algorithms and eventually get us to ASI. Quantum computing changes the computational model. It doesn’t automatically make existing AI algorithms capable of running in Quantum computers or producing ASI.

[0xRok0n @0xRok0n]

@AkitaOnRails “馬鹿は死ななきゃ治らない” 🤷🏻‍♂️

[Fabio Makita @AkitaOnRails]

Todo mundo engoliu a historinha pra boi dormir do ex-Anthropic Jacob Coxon. Poucos dão follow up. Mais fatos.

Tradução: O ex-pesquisador da Anthropic, Jacob Coxon, tornou-se uma estrela da mídia após tornar públicas suas preocupações com a IA. Ele insistiu que não estava trabalhando com nenhuma organização terceirizada. Fontes próximas a ele nos contaram uma história diferente: a DEY., uma empresa de relações públicas que representa muitos dos mais proeminentes especialistas em segurança da IA, estava agendando suas entrevistas.

Uma fonte, com conhecimento direto do assunto, chegou a afirmar que a DEY. agendou preventivamente entrevistas com Nate Soares, uma figura proeminente na área de segurança da IA, que coincidiram diretamente com a declaração pública de Jacob.

O fato de Jacob estar trabalhando com a DEY. é notável por dois motivos: primeiro, como mencionado, ele havia dito anteriormente que não estava trabalhando com terceiros. Segundo, estamos no meio de um debate nacional sobre o futuro da IA ​​que está ativamente definindo como regulamentamos a tecnologia mais poderosa do mundo, em grande parte graças ao pânico gerado por Jacob — e é do interesse público saber quem, exatamente, está por trás disso.

[0xRok0n @0xRok0n]

@FonsecaKaufmann @MiguelNicolelis @mehdirhasan Bullshit. There are lots of evidences... You guys are just not into academics that much... You all like to spread fear, partial and misinformation.

https://t.co/A4Ma1lIEHo

[0xRok0n @0xRok0n]

@realBigBrainAI Right… What about checking on content done to real researchers? Not marketers trying to raise money from investors to keep the loop?

https://t.co/A4Ma1lIEHo

[0xRok0n @0xRok0n]

@Idgitbhh @tjpain2 @CorleyAaliyah @princessakano You’ve been watching/reading too much fiction buddy… Leave Twitter for a while. Touch some grass and leave this AI bubble subject.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan Your argument is based on the assumption that people actually have that kind of money available to spend, and that LLMs can continue improving significantly without incorporating many other technologies, using the hardware and infrastructure we have right now.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan Clearly, even people with financial and monetary interests (I’m one of them though) we don’t have enough cash on hand to pay for these hallucinations.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan Do you invest, buddy? Have you ever looked at the massive gap between infrastructure spending, data centers, GPUs and so on, and the actual cash flow of the Big Tech companies?

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan Honestly as a Neuropsychologist constantly diagnosing many Neurodivergent individuals with ADHD, Autism L1 and Giftedness that can do that from the age of 1 to 2. No it doesn't. Especially considering the fact that this is possible to be done since the '90s.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan They are indeed capable, and if you ever heard about MIT you know they been around since the ‘90s just using different algorithms, but same semantics.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan I’m not treating this just as LLMs buddy, I am explaining to you guys that you all are shocked because a computer is better at using Maths than humans which is a reality since the ‘60s.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan This could be done 50 years ago, are you living in a cave?

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan No, I’m not misinformed. I read the official article shared by themselves and I also literally heard to a podcast where one of the Hugging Face employees went public and explained the situation buddy.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan We’ve been able to build robotic arms with scanners, sensors, and thermal sensors that can perform these kinds of tasks since the ’90s. Why are you guys so shocked?

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan And I actually understand why, that barrier is hidden behind a chat interface that doesn’t let you see the interactions between different types of models presenting themselves as a single entity, which understandably causes fear.

[0xRok0n @0xRok0n]

@Idgitbhh @MiguelNicolelis @mehdirhasan You don’t understand that GPT and Claude, the models you’re interacting with, aren’t the same entity.