Comment by kozikow | Hacker News Reader

Comment by kozikow | Hacker News Reader

Everyone is still left guessing what have happened.

At least my speculative view (rewritten by chatgpt ofc):

There's speculation that OpenAI's newly developed Q* algorithm marks a substantial advancement over earlier models, demonstrating capabilities like solving graduate-level mathematics. This led to discussions about whether it should be classified as Artificial General Intelligence (AGI), a designation that could have significant repercussions, such as potentially limiting Microsoft's access to the model. The board's accusation of "not consistently candid" against Sam Altman is thought to be connected to efforts to avoid this AGI classification for the model. While the Q* algorithm is acknowledged as highly advanced and impressive, it's believed that there's still a considerable journey ahead towards reaching a true singularity event. The complexity of defining AGI status in AI should be noted. At what point of advancement can a model be labeled as an AGI? We are way past simple measures like the Turing Test. Arguments from both perspectives have their merits.

I believe it's not possible to keep such thing a secret, so this reason as a main motivation for board actions sounds kind of weak to me

The progress from GPT3 to GPT4 has been so substantial that many might argue it signifies the advent of Artificial General Intelligence (AGI). The capabilities of GPT4 often elicit a sense of disbelief, making it hard to accept that it's merely generating the most likely text based on the preceding content.

Looking ahead, I anticipate that in the next 2-3 years, we won't witness a sudden, magical emergence of "AGI" or a singularity event. Instead, there will likely be ongoing debates surrounding the successive versions like GPT5, GPT6, and so on, about whether they truly represent AGI or not.

Correct, AGI refers to a level of AI development where the machine can understand, learn, and apply its intelligence to any intellectual task that a human can, a benchmark GPT-4 hasn't reached.

What actually happened between GPT3 and GPT4 was so called RLHF, which basically means fine tuning base model with more training data, but structured so it can learn instructions and that's all there was really + some more params to get better performance. Besides that making it multi modal (so basically sharing embeddings in the same latent space).

Making it solve graduate level math is a lot different than dropping some more training data at it. This would mean they changed the underlying architecture, which actually could be a breakthrough.

My guess this is just a good 'ol corporate power struggle.

There is no point in speculating until someone shows what the hell Q* even is.

Reply on news.ycombinator.com