Rumors mention recursive "self" improvement (training) already ongoing at big scale, better AIs training lesser AIs (still powerful), to became better AIs, and the cycle restarts. Maybe o1 and o3 are just the beginning of what was choosed to make available publicly (also the newer Sonnet).
https://www.thealgorithmicbridge.com/p/this-rumor-about-gpt-...
The pace of change is actually uncertain, you could have revolutionary advances maybe 4-7 times this year, because the tide has changed and massive hardware (only available to few players) isn't a stopper anymore given that algorithms, software is taking the lead as the main force advancing AI development (anyone in the planet with a brain could make a radical leap in AI tech, anytime going forward).
https://sakana.ai/transformer-squared/
Beside the rumors and relatively (still) low impact recent innovations, we have history: remember that the technology behind gpt-2 existed basically two years before they made it public, and the theory behind that technology existed maybe 4 years before getting anything close to something practical.
All the public information is just old news. If you want to know where's everything going, you should look to where's the money going and/or where are the best teams working (deepseek, others like novasky > sky-t1).
1. Positive rumors are profitable => they are targets for marketing activities, especially when huge money is at stake.
2. Humanity has a long history of false "fast technological success" rumors: thermonuclear fusion, a cryptocurrency that will disrupt the bank system, IoT that will revolutionize everything, AI boom at 80th, etc. They are almost always wrong.
3. Development cycles in IT are fast; on highly concurrent markets, they are extremely fast. The current public information in the AI industry describes almost the actual state of it. The risk "not to be the first one" is too high to hide or delay something. Such a delay may literally cost billions of investments.
Yes. The latest product releases from them all, have been chain of thought tweaks to existing models, rather than a new model entirely. Several models are perceivably the same or worse than previous models (sonnet 3.5 is sometimes worse than Opus 3.0 and Opus 3.5 is nowhere in sight.)
GPT4o is sometimes worse than base GPT4 when it was available.
The newest and largest models so far are either too expensive to run, and/or not much better than the previous best models, and so this is why they have not been released yet despite rumours that these newest models were being trained.
I would love announcements/data to the contrary.