Not really. Europe isn't going to miss out just because some things arrive here a few months later. The bigger issue is that people feel entitled to have everything immediately and think their own needs are more important than the collective good. I don't want to live in a world where nothing is regulated and corporations can abuse their users as much as they want. I don't want to have to worry about being poisoned by dangerous chemicals in the food I eat, and I don't want to live in a world where the most dangerous (and also one of the most useful) technologie is left unregulated
@kimmonismus People always say "Oh this is the year of agents or this is the year of humanoid robots" but I think what's even more important is that we are entering the year of science, next year so many math, physics, biology, chemistry and many more problems will be solved !
Yes I think so too, I also suspect that OpenAI has cracked RSI or is really close. They released Astra 54 days ago, they cut down on subscription value, and now they released 6.1 tera, which is on par with SOL but is cheaper, all that even though OpenAI should be the ones with the most compute, so I think they use all that compute internally
@kimmonismus I do think there are many small bubbles that are going to burst over time in the AI sector, this is one of them, another is the countless AI startups trying to just slap AI onto everything, but the big players are not going anywhere
I really hope they make it so that the new line of models is Luna, Terra, and Astra. For me, Sol was always in a weird spot; after Astra released, it was only a little better than Terra on most tasks but much more expensive and a lot dumber than Astra. We do not need more expensive models; of course, we want one for those who can afford it, but the rest need a reliable model that is intelligent and cheap, which for the 5.6 line of models was Terra. You could run it on Ultra, even back when I only had ChatGPT Plus; if you did the same thing with Sol, your weekly usage was gone in 30 minutes and nothing got done
@kimmonismus Terra was the only useful model when the 5.6 generation released. Luna is cheap but just too dumb, while Sol is too expensive for being only slightly smarter than Terra
@Dr_Singularity Well she is right tho, you may be able to reverse the effects of aging in the future, that does not change the fact that you are getting older, you cannot just say in 2100 "Well, my body is that of a 21 year old so I am 21" even though you were born in 2003 π
There is nothing wrong with being cautious, you keep saying you are a responsible optimist, but that is only true when you act upon it. How can you post about agents hacking websites and large scale cyberattacks, but as soon as you don't get the dopamine of a new SOTA model, you keep crying "me want better model now" ?
@kimmonismus It's simple the normal everyday user doesn't want to use the Grok bot, so there are far fewer restrictions than when the big players release something
@Thirumagal_12@kimmonismus You're right they can only do cybercrimes, completely harmless and luckily the models are the best they will ever be, so we shouldn't worry about security what a great take βοΈ
@Dr_Singularity It is not just bots, humans are the same way. I heard this quote once and it is so true: "I once had this misconception that by providing evidence and facts, you can change someone's beliefs"