
Dragos Roua
20.5K posts

Dragos Roua
@dragosroua
builder https://t.co/WFR15kJ8VH https://t.co/zsZU4i0Vl1 https://t.co/38Qtrx8KNa
Mostly inside myself Beigetreten Ocak 2009
1.9K Folgt5.6K Follower

@MarsDorian Yeap, that’s what you’re supposed to do. Tastes surprisingly good, though
English

@dragosroua Sugar?? Are you supposed to dip the fries into butter and then into sugar to make it stick? Doesn’t sound tasty tbh
English

Did you notice the question “How did Claude do?”. It appears more and more and now it seems to be mostly at the beginning of the session. And I think it dynamically adjust its effort for the rest of the session, based on your answer. For instance, if you answer 3, or good, it kinda plateau or gets number. If you give a 1 or bad, it intensifies somehow and puts more work into it.
So it’s not always the same Claude, it “follows” your expectations, just like Facebook adjusts its feed for your topics of interest.
English

@willreil I've caught it setting 10 minute timeouts on some things just for no reason, like a delay tactic.. weird.
English

The AI segregation has officially started. @OpenAI will share the latest model, GPT-5.4-Cyber, only with a “select few”, just like @AnthropicAI did with Claude Mythos.
There are now two AI tiers: first one is the rich and powerful - and second one is… the rest.
nytimes.com/2026/04/14/tec…
English

ngl, the Matrix theme of AIGernon is like a blast from the past. I can almost see a Beta Basic interpreter there. Almost.
aigernon.app

English

@DoctorYev Yes, it’s a relevant layer. But freedom is individual to each of them. Highly personal, cannot be outsourced.
Maintaining the state is a more predictable set of activities, and can be outsourced.
Just my 2 cents.
English

@dragosroua Sounds cool. Checking this out right away.
Followed you.
English

first: congrats!
second: that's how it starts. I mean, now, that we will have to define what consciousness is in order to evaluate if an AI is sentient or not, maybe, just maybe, we will finally understand consciousness.
Henry Shevlin@dioscuri
Big personal news: I’ve been recruited by Google DeepMind for a new Philosopher position (actual title), focusing on machine consciousness, human-AI relationships, and AGI readiness, starting in May. I’ll continue my research & teaching at Cambridge part-time. Absolutely stoked!
English


@gpsma2025 That's true, you only have yourself.
That's why it's important to develop yourself and be a nice person - you don't want to end up with an idiot in your late years.
English

The new hard skills are the old soft skills
Erik Voorhees@ErikVoorhees
The world doesn’t realize this yet but AI will finally make philosophy a highly marketable skill
English







