Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How will OpenAI compete?

ben-evans.com

Post reply on HN
Print view View original

371–380 of 693 posts

Re: How will OpenAI compete?

#371

Post by Gigachad »

Leynos wrote: post #366
Gigachad wrote: post #350

Earlier quoted context omitted.

At the conversation backlogs worth anything? To me they seem as valuable as Google search history. After maybe 3 days they are worthless.

People get attached to month long conversations, strangely. Sometimes even refusing to use the fork feature. And the memories are also something that adds to this greatly.

I guess if you treat it like a virtual boyfriend. Personally I found the memories to be an anti feature. I start chats to get a clean slate and test new ideas without previous ones polluting the chat.

Re: How will OpenAI compete?

#372

Post by XorNot »

wickedsight wrote: post #275
stingraycharles wrote: post #269

Earlier quoted context omitted.

Isn’t that exactly what’s being discussed re: OpenAI? They seemed unstoppable a few years ago, but have lost quite a bit of reputation and their position of technical lead.

> have lost quite a bit of reputation in the tech world, maybe. All my 'normie' friends are using ChatGPT though and have no concept of their reputation, nor intention of switching. Most people I know are hardly even aware of alternatives, even of Gemini, though everyone has a Google account. I personally also use ChatGPT and have zero reason not to, currently. I might switch if they royally mess up, but everything t…

But would they pay for it? That's the difference.

Re: How will OpenAI compete?

#373

Post by christophilus »

mgaunard wrote: post #312

As far as I can tell Google Gemini has the best overall integrations (Android, WearOS, Google Home) with the only voice recognition that actually works (Gemini Live). Anthropic Claude has the best integrations with coding; what would make sense is for them to focus on that segment. Other AI companies don't have anything really compelling. Meta has a model that's fully open-source, but then that's not particularly use…

I’m not sure. I started using Codex last week. Codex, at $20/mo is a very good value.

Re: How will OpenAI compete?

#374

Post by kavalg »

maxloh wrote: post #349
daxfohl wrote: post #130

I just wonder how long it'll take local models to be good enough for 99% of use cases. It seems like it has to happen sooner or later. My hunch is that in five years we'll look back and see current OpenAI as something like a 1970's VAX system. Once PCs could do most of what they could, nobody wanted a VAX anymore. I have a hard time imagining that all the big players today will survive that shift. (And if that partic…

In my experience with Gemini, most of its capabilities stem from web searching instead of something it has already "learned." Even if you could obtain the model weights and run them locally, the quality of the output would likely drop significantly without that live data. To really have local LLMs become "good enough for 99% of use cases," we are essentially dependent on Google's blessing to provide APIs for our loca…

Unless you can provide a (community) curated list of sources to search through (e.g. using MCP). Then I think local models may become really competitive.

Re: How will OpenAI compete?

#375

Post by amelius »

athrowaway3z wrote: post #347
xml wrote: post #295

Earlier quoted context omitted.

Even with inflated RAM prices, you can buy a Strix Halo Mini PC with 128GB unified memory right now for less than 2k. It will run gpt-oss-120b (59 GB) at an acceptable 45+ tokens per second: https://github.com/lhl/strix-halo-testing?tab=readme-ov-file... I also believe that it should eventually be possible to train a model with somewhat persistent mixture of experts, so you only have to load different experts every f…

I agree the parent is a bit too pessimistic, especially because we care about logical skills and context size more than remembering random factoids. But on a tangent, why do you believe in mixture of experts? Every thing I know about them makes me believe they're a dead-end architecturally.

> But on a tangent, why do you believe in mixture of experts

In a hardware inference approach you can do tens of thousands tokens per second and run your agents in a breadth first style. It is all very simply conceptually, and not more than a few years away.

Re: How will OpenAI compete?

#376

Post by epolanski »

shubhamjain wrote: post #29

Everyone is actually underestimating stickiness. The near billion users OpenAI has is actually a real moat and might translate into decent chunk of revenue. My wife, for example, uses ChatGPT on a daily basis, but has found no reason to try anything else. There are no network effects for sure, but people have hundreds and thousands on conversation on these apps that can't be easily moved elsewhere. Understandable tha…

I don't really see that stickiness to be honest.

Most people I know with android phones, myself included, just use Gemini which is bundled with the OS and has a dedicated button, has excellent data and integration with maps and such.

When it comes to enterprise, non IT companies (banking, insurance, etc) in Europe seem to be defaulting to Google's offerings, Gemini and NotebookLM in particular.

Re: How will OpenAI compete?

#377

Post by kavalg »

MagicMoonlight wrote: post #328
amelius wrote: post #299

Earlier quoted context omitted.

There will be companies producing ICs for cheap models, like Taalas or Axelera.ai today. These models will not be as good as the SOTA models, but because they are so fast, in a multi-agent approach with internet/database connectivity they can be as good as SOTA models, at least for the general public.

All they need to do is produce one for GPT-OSS and it’s over. That model is good enough for real uses.

I wonder why did they release it then.

Re: How will OpenAI compete?

#379

Post by y42 »

shubhamjain wrote: post #29

Everyone is actually underestimating stickiness. The near billion users OpenAI has is actually a real moat and might translate into decent chunk of revenue. My wife, for example, uses ChatGPT on a daily basis, but has found no reason to try anything else. There are no network effects for sure, but people have hundreds and thousands on conversation on these apps that can't be easily moved elsewhere. Understandable tha…

Why would you want to move conversations with you? I use multiple different models, I don't care about the history.

My "brain" in terms of projects, is local on my computer. I have a simple set of system rules that I need to copy.

I am not everyone, I understand that. What I try to say: don't overestimate the lock in effect of AI. I doubt there is one.

Re: How will OpenAI compete?

#380

Post by jmyeet »

shubhamjain wrote: post #29

Everyone is actually underestimating stickiness. The near billion users OpenAI has is actually a real moat and might translate into decent chunk of revenue. My wife, for example, uses ChatGPT on a daily basis, but has found no reason to try anything else. There are no network effects for sure, but people have hundreds and thousands on conversation on these apps that can't be easily moved elsewhere. Understandable tha…

I think that's false. The cost of switching is so low that the best product will win and there's no moat.

I honestly can't see how OpenAI can possibly recoup the hundreds of billions poured into it at this point. I'd say AI assistants are no more sticky than browsers or search engines.

You might be tempted to say that Chrome or Google are sticky. But they're really not. A lot of people aren't old enough to remember the 90s when we had multiple search engines and people did switch. I know this goes against prevailing HN dogma but I'm sorry: Google is simply the best search engine. It doesn't have a magical hold on people. People aren't fooling themselves.

And Chrome? Before smartphones it was simply the better browser. Firefox used to have a much larger market share and Chrome ate their lunch. By being a better browser. Chrome was I think the first browser, or at least the first major browser, to do one process per tab. I still remember Firefox hanging my entire browser when something went wrong. I switched to Chrome in version 2 for that reason.

And now browsers are more sticky because of Chrome on Android and Safari on iOS. Safari really needs to be cross-platform, like seriously so. I know they briefly tried on Windows but they didn't really mean it.

Anyway, back to the point. I believe there's a certain amount of brand inertia but that's it. If Gemini dominates ChatGPT performance and UI/UX, people will switch so fast.

Google, Microsoft and Meta can survive the AI collapse. Apple is irrelevant (at least for now). OpenAI? Doomed IMHO.

Post reply on HN