OpenAI and Ive

Fair play to Jony's career and influence at Apple, but in my mind the top designers at Apple are the folks who rescued their laptops from the dysfunctional mess Ive's thinness obsession put them in.

Also, the problem with this year's crop of AI accessories (which is all they are) has not been their industrial design. Get the bag, Jony, but let's not pretend this matters.

A revisionist history of Ive is well past due, now that we have some distance on it.

Ive's talent was to be a cipher for Jobs' taste and vision. ANY designer capable of instantiating Jobs ideas would have done just as well at creating a Rams-derivative design language for Apple.

6 years of the 2013 Mac Pro is Ive. Laptops too thin to adequately cool their GPUs is Ive. Cables that constantly split their strain relief, exposing the wires inside is Ive. Blinding white UI is Ive.

It's telling that the same people who are "interested" in Ive's OpenAI technological beanie baby, were openly critical of Humane's AI pin, given it's basically the same thing.

"any technology requiring hundreds of billions of dollars to prove itself is built upon bad architecture"

I'm not sure what exactly this means, but LLMs have clearly already proven themselves. Just look at all the writers who are suddenly out of jobs, all the engineers who now have these tools in their IDEs, all the machine translation tools that suddenly actually work.

"generative AI as a technology requires too much energy"

It only needs to be cheaper than humans, and I can run an actually good LLM on my MacBook Pro without the fans turning on.

"and training these models is equally untenable, both as a result of ongoing legal issues"

I think that's a bit of a misunderstanding of how laws work. Laws are only static for humans. They bend and break for companies. If you steal a book in a store, you get the cops running after you. If OpenAI steals all the world's books, they get to negotiate with the few other companies that hold actual power, and just keep everybody else's stuff.

"That’s interesting."

On a scale from original Bondi Blue iMac to Magic Mouse 2, exactly how interesting is this going to be?

@Someone

I wouldn't have thought it was Ive who made the decision to bet on OpenCL - he would have designed the Mac Pro 2013 to specs provided by engineers. I still love the Mac Pro as a design but yeah it wasn't future proof.

Ive would have designed the MacBook Air that the entire industry copies to this day - yeah not enough ports but was that Ive or engineers and bean-counters looking at ways to decrease costs and ensure healthy profit margins?

I don't know and I doubt anyone outside of Apple knows either.

@Plume

I think the quality of writing that chatGpt does is only good enough to be free of charge.

If they asked enough to be profitable few people would pay.

50$ for the most generic of outlines isn't an easy sell.

@Niall the OpenCL bet didn't necessitate any aspect of that machine's design, however. Everything about that computer was pure hubris (actually it was a deeply insecure statement - like a balding divorced middle aged man buying a sports car, if you want to truly psychoanalyse the company).

Apart from everything the fundamental design "feature"; the "Thermal Core" didn't work. It didn't adequately cool the machine, the same way his last thermal chimney design, the G4 Cube didn't adequately cool the machine.

Both exercises in aesthetics over utility, both with the goal of compactness, both with thermal chimney designs, and both cooked their own components; because they were *bad designs*. The same way his thin-obsession for laptops didn't adequately cool their GPUs. Decorative thin-ness, and decorative minimalism have marked his post-jobs tenure with Apple.

"I think the quality of writing that chatGpt does is only good enough to be free of charge."

Most people don't pay for things like news, so that doesn't matter. Writers (human ones) previously wrote the articles on all of these online sites, and now they don't anymore. What previously was a team of a dozen people is now one editor who goes through all the texts and makes them read like they were written by a person.

I meant that I don't think people using the LLM are willing to pay the actually cost of using the LLM.

As long as it's heavily subsidized by investors it works, but if that money dries up I think even the scammiest link farm will back off.

I might be too naive about this though, and as the saying goes: the market can stay irrational longer than I can stay solvent.

I think the large LLM companies are betting that the cost of running their models goes down faster than the investments dry up.

@plume “ all the machine translation tools that suddenly actually work.”

How can we know that for sure? Unless we speak fluently the 2 languages.

"How can we know that for sure? Unless we speak fluently the 2 languages."

You just answered your own question.

>Sam is insane. He managed to seal a chatgpt distribution deal with Apple while collaborating on an iPhone killer with Apple’s top designers.

Is that very different from Apple doing a Motorola deal for the ROKR while working on the iPhone?

Or Google doing an Apple deal for the iPhone while redesigning Android to be more like iPhoneOS?