I’m not sure how long ago that was, but LLM context sizes have grown exponentially in the past year, from 4k tokens to over a hundred k. That doesn’t necessarily affect the quality of the output, although you can’t expect it to summarize what it can’t hold on memory.
Gotcha. That sucks.
So those calls are not for the benefit of US companies?
Eh. Gen-x here. I still have an hour long phonecall over signal with my best friend over signal two times a week or so.
In my teens I wasn’t too happy about making phonecalls either, but working on a helpdesk for a while sure cured that.
On the other hand, I live in a country with consumer protection, so robocalls are not a thing. And I’d strike down upon thee with great vengeance and furious anger (and GDPR) those companies who attempt to poison and destroy my personal attention.
Honestly, between these obsession posts and all the other non-tech news that gets posted here, I just unsubscribed this morning. The signal to noise ratio in this community is just not worth it to me.
I prefer the latter, because it’s so much easier to filter out posts about Elon than it is to filter out posts about X (without creating a ton of false positives).
I might have been a teensie bit sarcastic when I wrote that ;)
Can you take your unopinionated headlines somewhere else? This is a technology community.
The sad state of political campaigning in 2024.
As long as ads and analytics are separate from each other and the rest.
Oh. Ok.
Why?
Good question. The answer is: for a significant amount of people, politics is emotional - so what makes sense isn’t necessarily relevant.
Before welcoming this as good news, be aware that democrats might also start thinking this misinformation is real, and decide to stay home and “not vote for a losing team”.
I guess responses like yours is the reason the headline didn’t mention the actual party gitlab is in talks with. People just love to have their villain.
Ignore the headline. Read the article. Gitlab is not about to sell to Google. They are about to sell to Datadog.
But they have been partially owned by Google for the past time, and the product has been great.
Google’s involvement is only going to lessen, so the only reason to put so much emphasis on that in the headline would be to get those rage clicks.
Typical that the title does mention Google (who currently has a minority stake) but not Datadog, who would become the new owner.
But yeah, I don’t foresee a new owner making things better for gitlab.
They could, but adding random zero width characters into words would also destroy ever spell checker, giving it away immediately and making sure that even unaware people would filter it. Doing it outside the words would leave them with too few spots to use for proper watermarking.
I think it’s far more likely they’ll use some kind of pattern in the tokens - that way the watermark will remain even when you don’t copypaste it.
But yeah, as said, they will never tell how it’s implemented, but it can still be simply subverted.
Yeah, no chance they’d rely on something that would be so easy to defeat. Watermarking by using word patterns is far more likely.
Still easy to defeat by just using another LLM to rephrase it though.
By that logic every news website is spam, because those also contain ads.
I agree the article is without much merit. But calling it spam because it also appears in a book and it mentions that source, is just diluting the term.
That’s a a bit too absolute way to look at it.
From their point of view the goal isn’t to abolish human involvement, but to minimise the cost. So if they can do the job at the same quality with a quarter of the personnel through AI assistance for less cost, obviously they’re gonna do that.
At the same time, just because humans having crappy jobs is the current way we solve the problem of people getting money, doesn’t mean we should keep on doing that. Basic income would be a much nicer solution for that, for example. Try to think a bit less conservatively.