Hacker News Reader

Comment by Havoc | Hacker News Reader

> Google dropped their biggest upgrade ever & ghibli core completely hijacked the zeitgeist.

On twitter maybe but on places like localllama google Gemini got way more airtime than this drama

Right — which is to say, the vast, vast majority of people heard about Ghibli and not Gemini.

The vast majority of people are consumers, not AI developers. Of course viral moments will be more consumer-oriented. It's easier to digest and reshare a Ghiblified-caricature than a research paper. But the content of that research paper will lead to the next viral moment years down the line.

I feel like you underestimate the degree to which AI developers are influenced in just the same ways as everybody else.

Capturing mindshare matters no matter how you do it. I would bet my life savings that more people will join OpenAI to work on their models as a result of the Ghibli moment than will join Google as a result of this particular (incredibly impressive!) Gemini iteration.

Because we’re humans! And we’re all pretty easily impressed, even those of us who actually build these dang things.

> I feel like you underestimate the degree to which AI developers are influenced in just the same ways as everybody else.

I feel like you're overestimating that. I work every day on developing LLM-based products, use them a lot while developing those, and have been doing so since November 2022 (beta release of GPT 3.5). So I should be a prime representative of the demographic you're talking about.

When a new frontier model is released, I run a series of tests, play around with it, and decide whether or not to use it based on that. Anything else would be ridiculously stupid, leaving either cost savings, product quality improvements, or development velocity increases on the table. We're not talking about databases or cloud VMs where there's a big switching cost. It's changing a model string and provider, so frictionless that it's a no-brainer.

We actually have some data that's on my side. Take a look at OpenRouter's model usage statistics. Note how OpenAI mmodels make up a tiny share on there, despite having more general mindshare on social media than all other model providers combined. Their data also shows people are very quick to switch when improvements to price/performance are released.

It's a bit like how vacuum cleaner specialists never use Dysons despite them having the most mindshare and overwhelming marketing budget. Regardless of vacuum cleaner specialists also just being normal people. Even if this one doesn't hold in your region, there's countless of these examples.

> It's a bit like how vacuum cleaner specialists never use Dysons despite them having the most mindshare and overwhelming marketing budget. Regardless of vacuum cleaner specialists also just being normal people.

Yes, but Dyson still crushes the market because it has the mindshare with non-specialists. That’s OPs point, uninformed people (the vast majority) will be impacted more by the “ghibli-core” memetic moment, and have a greater impact on future events than specialists who appreciate the Gemini 2.5 release

Back up the thread it was about the whole market, but then we started talking about AI developers (=the vacuum salesmen) and how much they're affected by the mindshare, which is also the part I quoted. The average consumer, absolutely, they all currently use GPT as the data shows, and are indeed led by Ghibli memes - though there's still plenty of time to change that by the huge players, á la Chrome eating all other browsers despite Microsoft's IE monopoly. AI developers though? Much less so.

the people who are qualified to work on the frontier of AI research are people with PhDs from leading universities who spend 80% of their day on arxiv reading CS papers, they don't learn about AI from Ghibli art on twitter. That's bread and circus for the masses. There's no correlation between who consumes that stuff and who builds the technology, if anything, the opposite. Leading edge chip designers at TSMC didn't end up there because they ended up playing graphic intensive video games

I just think you’re totally wrong about that, to be honest. It’s very strange to me to paint every AI researcher as an unfeeling robot who does nothing but read PDFs, who’s not remotely interested in how their work will ever be used by real people.

No, frontier model researchers are not learning about the concept of AI from Twitter.

But it’s one more piece of visibility, one more sense of “hey this company is actually doing stuff with their models,” that absolutely contributes to where these insanely in-demand people choose to do their work.

> Leading edge chip designers at TSMC didn't end up there because they ended up playing graphic intensive video games

What a baffling and wrong example! I mean, I guess I don’t know about TSMC specifically, but tons and tons of technologists have an origin story that rhymes with “I was really into computer games, then got curious how they worked”

This is my whole point: engineers are real people, and real people are fascinated by the uses of AI!

That's the people who can be hired and get paid big bucks right now. The comment is referencing people who might be inspired to start learning this stuff seriously.

Leading edge chip designers at TSMC didn't end up there because they ended up playing graphic intensive video games

Kind of a large assumption. I find it quite plausible that enjoying video games as a kid might lead to someone studying engineering as opposed to medicine or finance.

The people who work at frontier AI labs are sending around Ghiblified pictures of themselves like everyone else (except on Signal).

Chatgpt has 500mil users. Google with all it's distribution via Chrome, Android still probably doesn't have that many Gemini users

the fact that vast majority of people are consumers matter

Exactly. New image models are always exciting for two weeks or so. New LLM models are useful until the next,. better model comes out.

Which probably is just a small integer multiple of two weeks, but hey!

Reply on news.ycombinator.com