agiaccel

[agiaccel @agiaccel]

@thsottiaux

[agiaccel @agiaccel]

@LiveOverflow Whats the setup? Single gpt 5.6 sol in ultra?

[agiaccel @agiaccel]

@Kuuntelija Näinhän se menee, ei papereilla ole merkitystä. Ainoastaan osaaminen merkitsee.

[agiaccel @agiaccel]

@gdb "Perf per watt looking incredible." > How large improvement over existing chips?

[agiaccel @agiaccel]

@chetaslua do u think we will get gpt 5.6 this week or next?

[agiaccel @agiaccel]

@valigo models from where

[agiaccel @agiaccel]

@chetaslua @rileybrown @Raviteja1161 do you have early access to gpt 5.6?

[agiaccel @agiaccel]

@chetaslua @demishassabis 3.2 or 3.5 wdyt?

[agiaccel @agiaccel]

@chetaslua what do you know about gemini 3.5 hallucination rates?

[agiaccel @agiaccel]

@NVIDIAGeForce #007FirstLightRTX

[agiaccel @agiaccel]

@steipete Any advice for implementing feature with agent 1 and then verifying the work with agent 2? I think the human reviewer is bottleneck rn.

[agiaccel @agiaccel]

@slethion How

[agiaccel @agiaccel]

@MTSlive 15 years from what starting date?

[agiaccel @agiaccel]

@nicdunz Marketing

[agiaccel @agiaccel]

@mark_k @GoogleDeepMind When?

[agiaccel @agiaccel]

@gdb @RoundtableSpace Greg what do you think of closing the loop for agents playing fast paced video games? Would that be feasible at all? Maybe need to wait for local models to improve for that

[agiaccel @agiaccel]

@SebAaltonen How much more productive do you think AI tools make you right now compared to no AI tools at all? 2x? 3x?

[agiaccel @agiaccel]

Hype for GPT-5.5 is insane

[agiaccel @agiaccel]

@AiBattle_ U fr?