Créer un compteSe connecter

Recherches récentes

Aucune recherche récente

Options de recherche

Disponible uniquement lorsque vous êtes connecté.
mastouille.fr est l'un des nombreux serveurs Mastodon indépendants que vous pouvez utiliser pour participer au fédiverse.
Mastouille est une instance Mastodon durable, ouverte, et hébergée en France.

Administré par :

Statistiques du serveur :

583
comptes actifs

mastouille.fr: À proposAnnuaire des profilsPolitique de confidentialité

Mastodon: À proposTélécharger l’applicationRaccourcis clavierVoir le code sourcev4.3.4

#openai

102 messages72 participants0 message aujourd’hui
Public
AAKL

This story is from August 8 and was shared yesterday, It merits sharing again if only to illustrate the utter crap that new AI models represent while hoodwinking users into thinking they have found the fountain of eternal knowledge. I dread to think that a whole new generation of kids will be raised on AI's fabricated error-ridden "knowledge."

The Register: Meet President Willian H. Brusen from the great state of Onegon theregister.com/2025/08/08/gpt @theregister #ChatGPT #AI #OpenAI

@paco

The Register · Meet President Willian H. Brusen from the great state of OnegonPar Avram Piltch
Public *
Santiago

Altos ejecutivos de las BigTech se unen al ejército de los EEUU como tenientes generales, como antes los dueños de las fabricas de autos, ollas y metalúrgicas tambien se unian para fabricar balas, misiles, etc. Imaginate que van a fabricar estos kia... y la que se viene... #eeuu #usa #meta #openai #palantir #ejercito #usarmy #army #militar es.wired.com/articulos/por-que

WIRED · Por qué el ejército de EE UU nombró a ejecutivos de Meta, OpenAI y Palantir como oficiales de alto rangoPar Steven Levy
A répondu dans un fil de discussion
Public
Ramesh #NotGoingBack

⬆️ @counternotions

>> Routing prompts to the right #AI model is a difficult task that requires aligning an AI model to a user’s preferences, as well as the specific question they’re asking.

May be they should rebrand it as #ChatSycophant

>> The router then has to make a decision on which AI model to send the prompt to…

AI model picker for AI models sounds very much like "turtles all the way down"

#OpenAI #ChatGPT

techcrunch.com/2025/08/12/chat

TechCrunch · ChatGPT's model picker is back, and it's complicated | TechCrunchGPT-5 was supposed to get rid of all the complicated AI models to choose from in ChatGPT, but that's not how it's panned out.
Public
Miguel Afonso Caetano

"In the aftermath of GPT-5’s launch, it has become more difficult to take bombastic predictions about A.I. at face value, and the views of critics like Marcus seem increasingly moderate. Such voices argue that this technology is important, but not poised to drastically transform our lives. They challenge us to consider a different vision for the near-future—one in which A.I. might not get much better than this.

OpenAI didn’t want to wait nearly two and a half years to release GPT-5. According to The Information, by the spring of 2024, Altman was telling employees that their next major model, code-named Orion, would be significantly better than GPT-4. By the fall, however, it became clear that the results were disappointing. “While Orion’s performance ended up exceeding that of prior models,” The Information reported in November, “the increase in quality was far smaller compared with the jump between GPT-3 and GPT-4.”

Orion’s failure helped cement the creeping fear within the industry that the A.I. scaling law wasn’t a law after all. If building ever-bigger models was yielding diminishing returns, the tech companies would need a new strategy to strengthen their A.I. products. They soon settled on what could be described as “post-training improvements.” The leading large language models all go through a process called pre-training in which they essentially digest the entire internet to become smart. But it is also possible to refine models later, to help them better make use of the knowledge and abilities they have absorbed. One post-training technique is to apply a machine-learning tool, reinforcement learning, to teach a pre-trained model to behave better on specific types of tasks. Another enables a model to spend more computing time generating responses to demanding queries."

newyorker.com/culture/open-que

The New Yorker · What If A.I. Doesn’t Get Much Better Than This?Par Cal Newport
#AI#GenerativeAI#OpenAI
Public
Andrii Kuznietsov

🧠 Сем Альтман кидає виклик Ілону Маску, підтримуючи конкурента #Neuralink. Він став співзасновником нового стартапу #MergeLabs, що розробляє інтерфейс «мозок-комп’ютер». Також він займається залученням коштів, які, ймовірно, надходять переважно від венчурного підрозділу #OpenAI.

Public
Mela News :verified:

OpenAI riattiva vecchi modelli e offre più scelta agli utenti dopo il lancio di GPT-5. Ora si può scegliere tra:
🤖 Auto
⚡️ Fast (GPT-5 mini)
🧠 Thinking mini
🧐 Thinking
Sono tornati anche GPT-4o, 4.1 (1M token), o3, o4-mini.
#OpenAI #GPT5 #IA

ExplorerFlux en direct
À propos