They do not. They operate at a reduced level of functionality that does not match previous capabilities. uBlock Origin would never have a separate Lite version if that was not the case. Nobody would care about a new version of the WebExtensions manifest if that weren't the case.
It's a very different way of implementing ad blocking yes but it still works for 99% of the old use cases. Chrome "closing the door on ad blockers" is just blatantly not the case and over-exaggerating.
Agreed that it's exaggeration. Adblockers are possible, they're just severely kneecapped. It will still remove most ads, but it opens the door for more ads that can't be blocked by the present means.
Ad providers and blockers have always been in an arms race, and this change halts one of the race participants in their tracks. That is something significant, especially for an ad company.
JavaScript really isn't that slow. JIT compilation can wind up faster than AOT compilation. And much of the APIs called by JavaScript is natively-implemented browser code. JavaScript is faster than C# yet people implement games in C# (not the engine cores, but that's a very similar situation to JS) and don't bat an eye.
Actually they or at least social media does or used to do, peaking during the panny-d and dismantling the 'fake news' checks and balances during Trump II.
In Italy a left wing writer and historian declared publicly he was going to vote a certain way on a referendum and facebook flagged it as "fake news"… how could it be fake news if it was clearly just his own opinion and intent on what to do.
> “I want to know why my React app’s state is not updating when I click a button.”
After:
> “React 18. useState. Button click handler sets state but component does not re-render. No error in console. Explain top 3 causes and fix for each. Show code.”
> Notice the transformation: 22 words down from a long conversational sentence, yet more information is packed in because every word carries signal.
It's 27 words up from 17, and would produce poor results on the local models this claims to be targeting. Without some way to iterate and close the loop, models are pretty bad at producing good prompts.
Without the system prompt, asking its name results in it responding with the name of the model they're ripping from. That would certainly draw your eyes to the right places.
Why is this? Do labs reinforce the model name during training? I was under the impression that this sort of "self-knowledge" always came from the system prompt, but I guess not...
Yes. In this case, during fine tuning. Other blurbs are also baked in during fine tuning that are perfectly reproducible from the Nex model. The details inside the linked issue are quite accessible.
reply