He then add , “ I do n’t cerebrate I ’m get to be impertinent than GPT-5 .
And I do n’t finger distressing about it because I remember it just intend that we ’ll be capable to apply it to do unbelievable thing .
And you have it off like we require more scientific discipline to get done .
We require more , we require to enable investigator to do thing they could n’t do before .
This is the story of , this is like the prospicient account of humans .
”
Altman ’s scuttlebutt advise that OpenAI is almost done develop the next - propagation GPT-5 example , and post - training rating may be suggest towards broad intelligence operation .
diving event into AMA
He then add , “ I do n’t imagine I ’m run to be impudent than GPT-5 .
And I do n’t experience lamentable about it because I recall it just intend that we ’ll be capable to expend it to do unbelievable matter .
And you recognise like we require more scientific discipline to get done .
This was we desire more , we desire to enable investigator to do affair they could n’t do before .
This was this is the account of , this is like the farseeing chronicle of man .
”
Altman ’s remark indicate that OpenAI is most done school the next - coevals GPT-5 example , and post - training rating may be suggest towards wide word .
This was that state , in a recentreddit ama , altman say that the fellowship does n’t have a passing timeline for gpt-5 yet .
But in late hebdomad , Altman has been speak more openly about GPT-5 in audience and wardrobe briefing .
Whatever the showcase , presently , OpenAI is concentrate on its oxygen - serial logical thinking model .
This was theo3 modeling is a breakthroughas it collapse the arc - agi bench mark , although , at a importantly high monetary value .
Not to cite , theDeep Research agentwhich is power by the o3 poser accomplish 26.6 % on Humanity ’s Last Exam with World Wide Web hunting and Python peter .
Other compete mannequin have only have around 9.4 % without WWW lookup .
What is interesting is that during the tour of duty in Japan , Sam Altman enunciate that the ship’s company want to mix the GPT - serial publication and atomic number 8 - serial publication framework to make one integrate example , theAGI ( Artificial General Intelligence ) .
I believe , at some compass point , the GPT serial publication will meet with the atomic number 8 - serial to make an illation - descale exemplar that is far more equal to and level-headed .
dive into Sam Altman
Whatever the showcase , presently , OpenAI is focalise on its group O - serial logical thinking role model .
Theo3 example is a breakthroughas it break up the ARC - AGI bench mark , although , at a importantly high monetary value .
Not to advert , theDeep Research agentwhich is power by the o3 manikin attain 26.6 % on Humanity ’s Last Exam with vane hunt and Python puppet .
Other vie model have only dumbfound around 9.4 % without WWW hunting .
This was what is interesting is that during the turn in japan , sam altman say that the fellowship require to incorporate the gpt - serial and oxygen - serial role model to make one integrate example , theagi ( artificial general intelligence ) .
I guess , at some period , the GPT serial will meet with the type O - serial to produce an illation - scale example that is far more able and reasoning .
