@DogeDELUX
Personally, I see the appeal with this model. LLMs aren't just for coding or reasoning. I like it when LLMs tackle the human element too, because yes I'm one of those people who likes to blur the boundaries between AI and human. I do care about the vibes, and if they can make AI sound more realistic in terms of the subtleties that makes up organic human dialogue, then I am all on board with it. For now it is way too expensive, but I don't believe they won't lower down the prices soon. Just imagine, within a year or two this may become the standard OpenAI model.
@JC-xx5dm
Why are most of the people trashing it seemingly focusing on its reasoning and calculation ability when its been clearly stated that its a chat , emotional support , conversational AI model.
@verduona
Since the cut off is 2023, then the downstream models were published already...(?) and now this is the late public release of the foundation model(?) - makes more sense to me...
@rexmanigsaca398
4.5 is a conversational model that's why it has huge data. It is not a reasoning model and for coding. In GPT 5, they will combine all these model into one whole intellect. Combining the EQ and IQ.
@Tech_Enthusiast_001
Strange one. I don't see much use in it for the general public to be honest. It is just more expensive (much more) and very slow. What it offers is very little. It is basically worse for everything but for "sounding human and empathic". So... digital waifus?
@WINTERMUTE_AI
I think its awesome! I hope they can bring the price down.
@danielneu9136
honestly it seems like people don't really get how powerful a larger model (and thus better worldmodel and world understanding ) is. All models are good at coding and math nowadays, but even if that is at the base of almost all products, most value lies in those tasks that require the soft skills that GPT4.5 excels at. It may unlock a whole range of use cases in Marketing, Sales, psychology and so on (bc honestly the models we have are not good at that, they just generate loads of nonsense!) But if you look at how persuasive GPT 4.5 is in comparison to other models (Make me pay Benchmark) the value really comes across. this simply is not a coding or math model, but that does not make it less valuable.
@Kutsushita_yukino
atleast it gathers everyones attention to different kind of task other than reasoning. some model might come out of nowhere better than what this model is trying to achieve, with a better cost
@flrn84791
I'd qualify it as way too expensive, and way too slow. But yay, new openAI model 🙃🙃
@XeTute
ngl GPT4.5 is pretty bad.
@micbab-vg2mu
i do not see any place for this model in my workflows - even in creative writing - Claude and Gemini work better.
@complianceaves1120
Is this what "Sam Saw" last year that got the Open AI scared and Sam fired I wonder ?
@Guyverman01
Could GPT-5 be the first actual reasoning model?
@Wanderer2035
I think see the angle they’re trying to take. Which is that most people aren’t geniuses, people are just average people. It doesn’t hurt to create an AI that isn’t hyper focused on being a genius, but instead optimized to more being just like an average person. This makes more sense if you want AI to inter mingle with society as our friends, like cyberpunk, instead of just super genius scientists or over professional secretary assistants
@AnkitPuniaFX
Look at Nvidia stock now, LOL.
@ShivrajSingh-po2ys
First benefit goes to businesses that too larger ones
@hydrohasspoken6227
"most interesting"? That says a lot
@MelikeBalcı-n8i
Appreciate the detailed breakdown! Just a quick off-topic question: My OKX wallet holds some USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). Could you explain how to move them to Binance?
@UsmanAli-ve6tq
Why is this model so expensive and who will use this?
@aimaniaco