That tldr needs a tldr…
But also you absolutely can learn & build small versions of LLMs on a regular laptop. I did it on my old 2017 Dell XPS, and trained it on the complete works of Shakespeare. It learnt to write almost passable Shakespeare hallucination in a couple of hours. There’s a good tutorial online if you search for it.
A 99% reduction from $2.9 billion is $29 million. On the scale from $0 to $29 million, $7,386 is much closer to $0, so it’s more accurate to say the deal was cut by 100% unless you want to quote it with several decimal places.
It’s 99.9997% to 4 places.
And a lot less than $29,000,000.
How is $2.9bn to $7k a 99% cut? It’s far closer to a 100% cut…
This guy resists.
Can you make a leaderboard? We can see which of us is closest to LLM—I’d place myself quite high up.
The fact you haven’t seen it and that it’s so common in LLM output just means there’s a huge amount of the internet you don’t look at. That could be a good or bad thing—depending on your perspective.
Countries which support the implementation of Chat Control:
Spain, Romania, Portugal, Malta Lithuania, Hungary, Ireland, France, Denmark, Croatia, Cyprus, and Bulgaria.
Countries that are undecided:
Belgium, Greece, Italy, Latvia, Slovakia, and Sweden.
Countries which oppose Chat Control:
Slovenia, the Netherlands, Poland, Luxembourg, Germany, Estonia, Finland, the Czech Republic, and Austria
Not like the good old days.
Those IP addresses were about the most believable thing in that episode. It’s occasionally amusing nonsense, requires about 1% concentration, and if you fall asleep in one episode and wake up in the next, you won’t need to reach for the remote.
It’s from FUBAR. I think they took the show’s name as a general directive.
I’m replacing a trivial pcb in a battery pack for a small fan. It’s 6V (4xAA) with a SPST top left and a USB socket top right. But why on earth would the power line on the pcb follow that curved path? There are no other components, just the 2 ends of the battery pack and a plastic housing.
Maybe just ‘because you can’?
Oh sure. The training data is about 5mb so easily manageable on a relatively modern machine, but not on the kind of thing that was used for ELIZA.