@joerossi98 @thsottiaux @OpenAIDevs yeah, I don't think its that too. However, at some (I think nearby point) they will have to merge those, if they plan to release their superapp
English
Alex (ai & it-sec)
263 posts
Alex (ai & it-sec)
@CSxAI
software engineer / MSc student / research assistant i take matrices and add them
Germany Katılım Şubat 2025
361 Takip Edilen25 Takipçiler
@CSxAI @thsottiaux @OpenAIDevs I've heard the same, but it seems untrue. In my case I was only using Codex
English
Am I the only one seeing more Codex usage today? It seems like it's burning through everything
@thsottiaux @OpenAIDevs
English
@JustinBleuel @willdepue well, you cannot select all options. e.g. to subscribe to pro from free tier on iOS, you'd have to subscribe to some other plan and then upgrade. I think not many are using this way, but those who would want to do so are unable to do it
English
@maurerchr Außerdem reicht "KI in der Industrie einsetzen" nicht! Die EU braucht Frontier-Modelle die hier entwickelt werden!
Deutsch
Bin nicht sicher, ob Hightech und "Regierung will Milliarden lenken" wirklich zusammenpassen. Regierung sollte sich auf Rahmenbedingungen konzentrieren, in denen Zukunftstechnologien florieren können
Bundesministerium für Wirtschaft und Energie@BMWE_
Mit der Hightech-Agenda will die Regierung Milliarden in Zukunftstechnologien lenken. Ein Megathema ist KI. Im Interview mit der Rheinischen Post betont Ministerin Reiche: „Jetzt geht es darum, KI in der Industrie und im industriellen Mittelstand einzusetzen.“
Deutsch
@theo forking company
triangle company
what symbol will yours have?
English
@MParakhin It will likely happen when they merge ChatGPT and Codex into their "Superapp".
English
I strongly believe the day will come when Pro will be available in Codex (possibly at API rates). In that beautiful future we won't need to run Pi separately or copy-paste text between Web Pro and Codex. And Codex won't perform training by hardcoding results :-)
English
@Simon__Grimm Not a bad idea!
1. "good people running for government" requires a system where good people are promoted in political parties.
2. "Parliamentarians should get decent salaries": Yes but where is the end? Really good (e.g. AI researchers) are payed hundreds of millions?
English
All parties except the CDU are currently discussing freezing MdB's pay this year. It's a bad, populist idea.
Parliamentarians should get decent salaries: it's a tough job and we need good people running for government. And the fiscal impact is tiny!
Luis Garicano 🇪🇺🇺🇦@lugaricano
Countries get the cabinets they pay for. Singapore pays its Foreign Minister about S$1.1m, around US$800,000. The salary is benchmarked to 60% of the median income of the top 1,000 Singaporean earners. That is why you can get Vivian Balakrishnan, former eye surgeon and hospital chief executive, implementing @karpathy's external brain idea (link below). The speech shows deep understanding of AI and fills one with confidence about Singapore's future. The UK Foreign Secretary earns roughly £165,000: the MP salary plus a ministerial salary of about £67,000. The ministerial part is frozen since the crisis and is down by roughly a third in real terms since 2010. This is what a junior Magic Circle lawyer earns. Spain pays its ministers around €85,000. So you do not get a surgeon who has run hospitals. You get a party loyalist who has never run anything.
English
@MonikaSchnitzer @privitera_ @handelsblatt @ki Rechenzentren alleine reichen nicht! Es braucht Firmen die hier vergleichbare Modelle entwickeln! Ansonsten kann der Zugriff auf solche Modelle trotzdem eingeschränkt sein.
Deutsch
Ein Beitrag von @privitera_ und mir im @handelsblatt :
Deutschland muss den Zugang zu Spitzen-KI sichern:
Neue @KI -Modelle wie #Mythos zeigen, wie gefährdet Europa durch Technologie aus den USA ist. Dagegen hilft nur eins: Infrastruktur auf eigenem Boden fördern. Warum?
Deutsch
1.3 million dollar in tokens in 30 days 😱
Peter Steinberger 🦞@steipete
The latest CodexBar update renders API costs wayyyy nicer. codex.bar
English
Alex (ai & it-sec) retweetledi
amazing post and great timing w.r.t. ant's post yesterday
we must build open ai to not get locked in by the vendors who will decide who gets which capabilities
and the west has to realize that open models are important and support open model efforts (like @arcee_ai, @NVIDIAAI)
Bill Gurley@bgurley
A new @bgurley blog post! I have been thinking about how sophisticated executives are using open source in super creative ways. Started writing this three years ago. Excited to finish it up and publish it! And with the new @p3institute brand. substack.com/home/post/p-19…
English
How is no one mentioning the obvious thing? Why can I not select a model in pi that requires me to play the role of the LLM too.
Lets think ourselves again! Its way more token efficient too. Also no more running into usage limits!
Mario Zechner@badlogicgames
sneak peak of the new pi.dev no tools mode. you're gonna love it. you can try it with `pi -nbt`. enjoy!
English
Alex (ai & it-sec) retweetledi
The quantity of code that devs ship has roughly 10xed. But net developer productivity (value created by unit of time) is only up by a bit, if at all. Part of it is that the additional code is solving more incremental problems. A bigger part is that the new code is creating problems of its own.
English
Alex (ai & it-sec) retweetledi
Anthropic is anticipating another big step-function increase in AI model capabilities in 2028 which will come with Nvidia's Rubin Ultra NVL576
I personally think there will be again issues with cooling and powering these insane 600kW racks. So with some delays factored in, late 2028 or early 2029 seems like a good guess for the arrival of 100T models.
English
@giffmana maybe marketing people are all seeing sparks of AGI...
English
go touch some grass and you can now actually close your laptop
Andrew Ambrosino@ajambrosino
what did they mean by this
English
@ajambrosino go touch some grass and you can now actually close your laptop
English
Which is good no? It should know how to retrieve factually correct data from the internet, how to filter data, how to rate information related to a question. Just like a competent human!
Flowers ☾@flowersslop
Frontier models do NOT already know most public knowledge. Even GPT-4.5 probably knows less than idk 70%. I dont even mean semi private facts, I mean ordinary public domain knowledge, like who played this minor character in this episode of this Belgian TV show form the 80s
English
@flowersslop Yeah, wondered about Sam's post related to developers as well.
My guess: being the nice company in public opinion actually leads to a LOT more revenue (see when Anthropic initially declined a deal with the government and celebrities posted ❤️ for that).
English
Why did OpenAIs messaging shift from ASI God to ASI tool
1 They hit a major wall
2 They believe in it but softened the messaging to avoid panic
3 They genuinely shifted toward seeing AI as augmentation
4 AI as augmentation is a more culturally sustainable framing internally
English
@scaling01 @FakePsyho But its great for PR! At what point would it be ASI or AGI though? 240? 300?
English
yeah im a bit confused about this one too
what's the point of trying to force the IQ scale onto model intelligence when models are improving so much every month. I also think it creates a lot of confusion because I don't see how this maps to real human IQ scores. It just seems like an arbitrary axis that looks IQ-like but is not really comparable
English
protip: if you see AI and IQ in the same sentence, you can be sure it's just pure webslop cosplaying as science
Ryan Shea@ryaneshea
Today I’m launching AI IQ — frontier AI models, scored on the human IQ scale. Instead of endless leaderboard tables, AI IQ shows: • Where models land on the IQ bell curve • How frontier IQ is changing over time • How models compare on IQ and EQ • What intelligence costs in practice GPT-5.5, Claude Opus 4.7, Gemini 3.1, Grok 4.3, Kimi K2.6, Qwen3.6, DeepSeek V4, Muse Spark, and more. Link in the first reply. Curious which chart surprises you most.
English