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

Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

openai.com

Post reply on HN
Print view View original

131–140 of 290 posts

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#131

Post by judge2020 »

1saadcodes wrote: post #128

I dont get why they'd make this available to free users when they're probably dealing with compute constraints considering their competition with the Chinese models. Giving a more capable model to a huge free user base looks like an expensive choice to me

IMO Competition with chinese models frees up higher-turn and bigger-context tasks, since SWEs are more likely to use model routers for coding. Thinking may be a few more steps but the problems are a lot less complex and probably 1/10th the context size, even when skills and memory are at play.

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#132

Post by orbital-decay »

ilaksh wrote: post #23

> Our mission is to ensure that artificial general intelligence benefits all of humanity. We’re introducing updates to ChatGPT that improve everyday conversations while expanding access for Free users. This clearly implies that they believe ChatGPT models are AGI and are now willing to say it out loud. Which I think is a fair interpretation of the term. They are general purpose intelligence in that you can get help f…

It doesn't fit their own definition of AGI, although it's inevitably vague [0]:

>artificial general intelligence (AGI)—by which we mean highly autonomous systems that outperform humans at most economically valuable work

[0] https://openai.com/charter/

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#133

Post by iJohnDoe »

> For Plus and Pro users, we’re updating GPT‑5.6 Sol in Chat to be more reliable with facts and provide more focused answers.

I was not impressed with 5.6 and this hits exactly why.

Also, this instant, medium, and high slider situation we now have everywhere is batshit crazy. It’s a major step back in technology.

I’ll put money on the table there will be a surprise in revenue because users have no clue what to choose, so they are constantly choosing high because they don’t want to risk getting inaccurate answers. If this was intentional by the dark patterns department, then brilliant. However, I’m guessing Anthropic and OpenAI are struggling to know how to deploy their models.

Also, the models were already amazing. They need to slow down and do a model release once a year and only do extremely minor iterations instead. Some amazing things are accomplished, but how people actually want to utilize the models gets screwed up every time in the process.

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#135

Post by aryehof »

For me on a paid plan, the effort indicator was hidden and the model was 5.5 instant. I had to press the + button to select “think harder” before the dial that allowed Sol medium or high to be selected to show.

It made me wonder how many paid subscribers realize they are using the same 5.5 instant model as free users by default. A dark pattern or oversight?

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#136

Post by willy_k »

ekidd wrote: post #127
johnsmith1840 wrote: post #110

Earlier quoted context omitted.

I typed wrong I mean I don't see how they will beat google. Google is the only true vertically integrated AI company meaning they own price at the bottom margins with their own chips and models made just for them.

Google's public AI model lineup is pretty bad right now. Gemini 3.1 Pro is scoring worse than some mid-sized Chinese models at 1/20th the cost, and their Flash and Flash Light models are horrendously overpriced on task benchmarks compared to GPT 5.6 Luna or the mid-sized Chinese models. There's no reason why Google's public stuff is this stale, overpriced and underwhelming. But at least until their next round of mode…

That doesn’t matter very much for the free tier market though. The game is to have something useable that is actually used, and that can be offered at sustainable margins. Google has significant advantages in those regards.

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#137

Post by red_green_yell »

keeda wrote: post #111
stymaar wrote: post #71

Earlier quoted context omitted.

Ads work when you can server $.01 worth of ads to a user for $.0001 of server cost. I fail to see how you can make it work when doing LLM inference which is significantly costlier than web search.

That's a problem Google has to solve too ;-) To expand: it seems inevitable that Google's SERP format will be replaced with a conversational / chatbot / agentic interface, which equalizes the playing field for all chatbot providers. This is because you can stuff in much fewer ads into a chat interface compared to SERPs. (They could try stuffing more ads but that would likely just push users more to the competition wh…

Google has multiple massive cost advantages over OAI: TPUs, free access to massively valuable data (search index, gmail, maps reviews/PO/ navigation, youtube, android etc.), lower talent costs, lower training costs, lower distribution costs (they can shove AI down our throats in so many different places), lower ad infrastructure costs (it will be a lot of work for OAI to recreate adwords), lower ad sales costs (OAI deploying an ad sales team will be a massive investment).

There's no way OAI has a long term advantage over google in replacing the search engine experience.

Google's one major weakness is it will face the innovator's dilemma as their core search revenue gets cannibalized. But they seem to have been able to get their entire org to recognize that AI is an existential threat so at least that's a good sign.

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#138

Post by willy_k »

keeda wrote: post #111
stymaar wrote: post #71

Earlier quoted context omitted.

Ads work when you can server $.01 worth of ads to a user for $.0001 of server cost. I fail to see how you can make it work when doing LLM inference which is significantly costlier than web search.

That's a problem Google has to solve too ;-) To expand: it seems inevitable that Google's SERP format will be replaced with a conversational / chatbot / agentic interface, which equalizes the playing field for all chatbot providers. This is because you can stuff in much fewer ads into a chat interface compared to SERPs. (They could try stuffing more ads but that would likely just push users more to the competition wh…

Have you used the Google search engine in the past couple years?

You’re painting a dichotomy that doesn’t exist and hasn’t for a good while.

Edit: I missed your final paragraph. I still disagree with this argument though, people still want to go to websites.

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#139

Post by Marha01 »

Jtarii wrote: post #38
redox99 wrote: post #26

Earlier quoted context omitted.

Because the model can't read your mind and know if you want a quick answer, or an hour long deep dive.

Then it should just ask the user what they want if its unclear from the context.

How is asking better than a reasoning slider?

Re: Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users

#140

Post by eru »

ekidd wrote: post #127
johnsmith1840 wrote: post #110

Earlier quoted context omitted.

I typed wrong I mean I don't see how they will beat google. Google is the only true vertically integrated AI company meaning they own price at the bottom margins with their own chips and models made just for them.

Google's public AI model lineup is pretty bad right now. Gemini 3.1 Pro is scoring worse than some mid-sized Chinese models at 1/20th the cost, and their Flash and Flash Light models are horrendously overpriced on task benchmarks compared to GPT 5.6 Luna or the mid-sized Chinese models. There's no reason why Google's public stuff is this stale, overpriced and underwhelming. But at least until their next round of mode…

It's interesting that Google fell behind before and managed to get back up to the frontier. But, boy, is their performance in this race uneven.

Especially considering that in the 2010s they were The Big AI company, especially after buying out boutique shops like DeepMind.

Post reply on HN