What's your take on the call to pause GPT-4.5/GPT-5?

They want it to pause so they can work on their own. I doubt their concerns are sincere just business as usual.

Strongly disagree Feels like a way to stifle the competition

it's great marketing.

disagree!

disagree, why 6 months pause 🫠

Are some companies fighting for extra six months to catch up with OpenAI? Yes! Clearly. Is It Unfair? Hmmm... Google R&D guys made a gift of a lifetime for OpenAI. They opened research empowering the IT community to make great products around it. What OpenAI did?! - pretended to be a non-profit organization, and then BOOM - commercialized it via Google's competitor #1. They hide technical details of the latest release, started fight-race with Google Search for monopoly, launched their own moderated "App Store" (soon collecting tribute from all authors), and implemented "sponsored true" in the search engines claiming OpenAI is just an LLM, while clearly, it's content aggregator... Yes. These papers are not about the danger of AGI. It's about community culture that pushes innovation forward and opportunists who exploit the openness of the community to dominate the market.

This call to action isn't completely clear. What's the objective really? Who are the right people/institutions to do this pause? What can the 6 months timeline do? Other companies/countries are developing their own, do they have to pause as well or is it just Open AI/USA?

Let's say I would open a store where you could buy everything with no limits for just 20 USD / month. I think we could agree - that this by itself would not be innovative. But still everyone would think it was innovative and would want to buy stuff there just to avoid paying "fair" prices elsewhere. Also no one would care where these products came from - and why my store can sell them for just 20/month. Maybe I stole the products from other stores. Maybe I bought them from factories with absoluetly no social and ehtical commitments. Or maybe my store is full with counterfeits. As long as no one asks - theres no problem. The AI companies are doing just that - stealing content from other "stores" and selling them for human unworthy prices. They try to hide that fact by focusing on the "innovation" Part. I agree, that the idea of interacting with the "web" via chat maybe somewhat of an innovation - but we are all just blinded by the absoluetly unrealistic prices for which we can get content now. If midjourney had to compensate the creators of the content from which midjourney "learns" - the prices of a midjourney subscription would be way higher. So yes - we need a short break, to talk about that. I'm on producthant every day - and I love innovation. But at the moment many of the AI tools are not innovations - but just an example of the ruthlessness of some entrerpreneurs.

I do agree. Unfortunately AI at this scale is too powerful of a tool to run it without guard rails. If we let capitalism get it's dirty hands on it we'll definitely end up with Skynet. Belee dat.

While large language models may have both positive effects (increased productivity) and negative effects (climate effects, proliferation of unsourced media) and wide-ranging economic impacts, the letter doesn't really propose any concrete actions on how to address any of these concerns, short of an indeterminate pause. Also, bigger models exclusively mean better models. Part of solution will be new models that have more validation and safety checks built in, and to find new ways more efficient methods of training and evaluating these models.

100% disagree. It's like oldschool sellers protest on E-commerce.

Some business owners are signing this for the benefits of themselves and others who have real fear about what will happen if this keep going

Tech alarmists, disengenuous at best, ulterior motives at worst.

We should make sure it progress faster so we can have cheaper way of doing stuff instead of paying people to do it...

I think

Six months moratorium sounds suspiciously specific a pause, does it not? Sounds the Horse Trade when they saw Henry Ford getting it done. Sorry, ain’t happening.

Strongly disagree. It feels dishonest and deliberately playing to irrational fears to pretend that a sophisticated word prediction model is somehow a threat. These AI tools aren't being built to replicate human emotion, agency, or autonomy. They predict the next word in a sequence based on probability after consuming an enormous dataset. Musk and the others are trying to slow that progress by fearmongering and assuming the average person doesn't understand the basis of this AI explosion.

I think this is ridiculous ! Especially when you see which people are asking for it !

disagree