Suivre
- X@merettm
- LinkedIn@jakub-pachocki
- Wikipédiaen.wikipedia.org
Jakub Pachocki
Directeur scientifique d’OpenAI, qui a écrit en septembre 2026 qu’aucun laboratoire n’a résolu l’alignement et la surveillance au point de pouvoir continuer longtemps à monter en échelle à pleine vitesse de façon responsable, qu’il s’attend et espère voir des ralentissements volontaires se généraliser jusqu’à l’établissement de seuils de sécurité communs, et que la coordination internationale sur le développement de l’IA doit devenir une priorité majeure des gouvernements.
Fonctions
- 2021 – mai 2024Director of Research, OpenAIJoined OpenAI in 2017; as Director of Research led the development of GPT-4 and OpenAI Five (the 2021 start is Wikipedia's date; OpenAI's own announcement dates only his research leadership, to 2017).Source →
- 14 mai 2024 – aujourd’huiChief Scientist, OpenAISucceeded Ilya Sutskever, announced by Sam Altman on the day Sutskever's departure was made public.Source →
Au registre
6 sept. 2026 · Publication · OpenAI
We may be used to thinking of AI as tools, but some agents will be pursuing their own objectives. They will find ways to collaborate with people, by bargaining with, tricking or blackmailing them.
Chief Scientist, OpenAI à l’époque
From the section on scalable defense, which also says the idea of racing forward at all costs seems absurd once one internalizes the seriousness of the stakes.
6 sept. 2026 · Publication · OpenAI
This is a time that calls for extreme caution. I am concerned no one is prepared for the consequences of a continued rapid rise in machine intelligence.
Chief Scientist, OpenAI à l’époque
Written directly after saying that internal results give him a strong expectation that the current speed of progress could be sustained into recursive self-improvement.
6 sept. 2026 · Publication · OpenAI
Currently I believe that no lab has solved alignment and monitoring to a sufficient degree to continue responsibly scaling at maximum speed for much longer. I expect and hope for voluntary slowdowns to become commonplace until shared safety bars are established.
Chief Scientist, OpenAI à l’époque
The essay's closing paragraph, which goes on to call for international coordination on future AI development to become a top priority for governments.