First western headline that doesn’t mince

🤷 I dropped my previous phone with no screen protector, screen cracked. I dropped my current phone, screen protector cracked, screen is perfectly fine.

n=2, but enough for me to just keep a screen protector on at all times.

I’m not sure I buy that. They’re like bumpers on a car. They’re meant to crack to take the pressure off the display. I’d rather crack a $20 screen protector than the $200 display.

That is an impressive demo on their site! Definitely what I’ve been looking for thank you!!

Thanks for the rec. I’ve used their skins before didn’t realize they had screen protectors as well!

I’ve already chipped my screen and am in general pretty clumsy so I can’t go without.

Gotcha. All I can add is a vote for Plausible. It’s great!

I spun up Plausible for my company site. Pretty straight forward honestly. It’s completely dockerized so there’s not a lot going on.

Your biggest problem is exposing the service on your home server to the internet. I personally wouldn’t recommend it, but if you know what you’re doing it’s possible.

I see your point. Rereading the OP, it looks like I jumped to a conclusion about LLMs and not AI in general.

My takeaway still stands for LLMs. These models have gotten huge with little net gain on each increase. But a Moore’s Law equivalent should apply to context sizes. That has a long way to go.

What drastically better results are you thinking of?

We’ve reached far beyond practical necessity in model sizes for Moore’s Law to apply there. That is, model sizes have become so huge that they are performing at 99% of the capability they ever will be able to.

Context size however, has a lot farther to go. You can think of context size as “working memory” where model sizes are more akin to “long term memory”. The larger the context size, the more a model is able to understand beyond the scope of it’s original model training in one go.

Image file format with excellent compression. It’s designed for web browsers, so what you’re probably running into is compatibility with other programs. It’s fairly easy to convert though to GIF or JPEG formats though.

Now he’s going to pay the bots! 🤣

Views basically. The number of times your tweet gets shown in someone’s feed.

Tweet impressions. Still pretty high though.