They just didn’t all have instantaneous global reach while sitting on the crapper, like Henry Ford with a Twitter account.

Even the Hearst type ones usually had time and human machinery between the formation of idiotic notions and it reaching other people.

Only kidding. Kinda.

I’m simply not built for the “puzzle based memory trial in front of a live firing squad”. It will not measure my ability.

Gross house means almost certainly not maintained.

The king of AI isn’t the shop with the biggest/bestest model, or whose machine has the bandwidth and unified memory.

It’s the company who has the trust of a massive number of users.

Apple was smart enough to give OpenAI the finger on deep integration, and focus on building out the part that everyone really wants.

Whoever provides the llm inference, do it with the privacy protection, and with access to the personal data you want AI to have access to, but without sacrificing the former. They’re literally the only company on earth in a position to do this.

We’ll see if they’re going to pull it off, very soon. If they have, they’ll have managed to negotiate and capitalize on this crazy hype cycle better than anyone, by far.

They could become the trusted gatekeeper of AI services, doing the one thing that none of the rest can ever do… and have done it without lighting trillions of dollars on fire fighting an unwinnable war.

But also the unified memory on macs is compelling for somewhat related reasons.

If it weren’t for the supply chain nightmares on top of nightmares, we’d be experiencing a nice little moment.

I had assumed they’re just reselling inference, where they’re “routing money” the way Walmart is “routing money” from my wallet to whoever makes Cheez-Its.

Are they doing something unique from resellers there?

The half usage limit still applies.