Sl00k

which does support the idea that there is a limit to how good they can get.

I absolutely agree, im not necessarily one to say LLMs will become this incredible general intelligence level AIs. I'm really just disagreeing with people's negative sentiment about them becoming worse / scams is not true at the moment.

I doesn't prove it either: as I said, 2 data points aren't enough to derive a curve

Yeah only reason I didn't include more is because it's a pain in the ass pulling together multiple research papers / results over the span of GPT 2, 3, 3.5, 4, 01 etc.

That's definitely valid, but just because a tool is used for scam doesn't inherently mean it's a scam. I don't call the cellphone a scam because most my calls are.

The jump from GPT-4o -> o1 (preview not full release) was a 20% cumulative knowledge jump. If that's not an improvement in accuracy I'm not sure what is.

Compare the GPT increase from their V2 GPT4o model to their reasoning o1 preview model. The jumps from last years GPT 3.5 -> GPT 4 were also quite large. Secondly if you want to take OpenAI's own research into account that's in the second image.

Curious why your perspective is they're are more of a scam when by all metrics they've only improved in accuracy?

Nvidia shield is an option you should check out.

Personally really been enjoying Kagi for the past year.

Perhaps a bad example because most people undermine them, but China has still decided to move forward with 4 different nuclear facilities this year despite having an ABUNDANCE of solar manufacturing. If they found that decision worthwhile I would think the opposite, assuming most of the reasoning is current battery tech can't sustain dark periods at a massive scale, but I'm not an expert.

Also just saw you mentioned nuclear costs in another comment, I suggest you look at South Korea and China's cost per facility compared to the US, they're able to build and maintain facilities at about half the US does.

If you have to ask that question you definitely don't use Tiktok it's far far superior algorithmically than Reels and YT Shorts which are both absolute garbage.

The article talks about why they'd prefer to shut down if you take their word it. Essentially the US is such a tiny portion of ByteDances revenue, it would be more optimal to shut down then to risk the sale of their algorithm. Assuming they're using relatively similar algorithms on Douyin, and they don't want whoever they sell to to turn around and sell to their Chinese competition, which is where the real money is being made for ByteDance.

Are US kids’ already dwindling attention spans going to be saved from exposure to the TikTok algorithm? Yes.

You're pinning the blame on tiktok when this also applies to YouTube (shorts and not), Instagram (Reels), Twitter. If we wanted an actual solution here we would implement actual children screen time laws, ironically similar to the under 18 gaming laws that have been implemented in China.

Tiktok is the only platform I've seen legitimate progressive movement on various issues and discussions centering on what that means and takes, in a way that actually fosters a great democratic progressive movement in the US.

From all I've read on this issue, not a single person has provided me with any insight into what or who this benefits that does not also apply to every other social media other than an entirely fabricated myth that they're controlling the algorithms to spread anti US sentiment. Anti-US sentiment definitely exists, but it exists as a discussion around what the US is currently doing. I.e. funding Israel, and as a counterargument to that I am also fed state department interviews on my FYP.

In the same vein Bloomberg just did a great study on ChatGPT 3.5 ranking resumes and it had an extremely noticeable bias of ranking black names lower than the average and Asian/white names far higher despite similar qualifications.

Archive source: https://archive.is/MrZIm