6 ms·

The GPT-5 Launch Was Concerning

86 points by csmeyer 1y ago

mutkach 1y ago
The focus now is not the model, but the Product - "here we improve the usuability by removing the choice between models", "here is a better voice for tts", "here is a nice interface for previewing html" Only about 5 minutes of the whole presentation are dedicated to enterprise usage (COO in an interview sort of indirectly confirms that haven't figured it out yet). And they are cutting the costs already (opaque routing between models for non-API users is a clear sign of that). The term "AGI" is dropped, no more exponential scaling bullshit - just incremental changes over the time and only over select few domains. Actually it is a more welcoming sign and not concerning at all that this technology matures and crystallizes around this point. We will charitably forget and forgive all the insane claims made by Sam Altman in the previous years. He can also forget about cutting ties with Microsoft for that same reason.
davydm 1y ago
Issues like this are why I don't use ai agents for code. I don't want to sift through the bullshit confidently spewed out by the model. It doesn't understand anything. It can't possibly "understand my codebase". It can only predict tokens, and it can only be useful if the pattern has been seen before. Even then, it will product buggy replicas, which I've pointed out during demos. I disabled the ai helpers in my IDEs because the slop the produce is not high quality code, often wrong, often misses what I wanted to achieve, often subtly buggy. I don't have the patience to deal with that, and I don't want to waste the time on it. Time is another aspect of this conversation, with people claiming time wins, but the data not backing it up, possibly due to a number of factors intrinsic to our squishy evolved brains. If you're interested, go find gurwinder's article on social media and time - I think the same forces are at work in the ai-faithful.
- mrits 1y ago
  There is a threshold that every developer needs for them to make it be worth their time. For me that has already been met. Your comment makes me think that you don't believe it will start producing higher quality code than you anytime soon. I think most of us are in the camp that even though we don't need AI right now we believe we will not be valuable in the near future without being highly proficient with the tooling.
  - bluefirebrand 1y ago
    > even though we don't need AI right now we believe we will not be valuable in the near future This reads to me like you don't think you're valuable right now either
cwrichardkim 1y ago
> They admitted that they were, and I am not lying about this, paywalling chat colors. […] This is a feature that a company adds when they are out of ideas This observation + sherlocking cursor suggests that perhaps sherlocking is the ideation strategy. Curious to see if they’re subsidizing token costs specifically to farm and Sherlock ideas
- dentemple 1y ago
  Yeah, I agree with the OP here. After all this time, being able to change the chat colors at this point has some real We-reached-the-bottom-of-the-backlog energy, and they're just now implementing the ideas that weren't considered important enough before by the PMs to consider. It hardly feels like a next generation release. As a related anecdote (not saying that this is industry standard, just pointing out my own experience), the startup I work for launched their app four years ago, and, for all four of those years, we've had "Implement a Dark Mode design" sitting at the bottom of our own backlog. Higher priority feature requests are always pre-empting it.
  - msabalau 1y ago
    The core product failure here is overhyping incremental improvement, eroding trust. PMs operating at this level ought to be bringing in some low cost UX improvements alongside major features. That simply isn't a sign that they've run ought of backlog. (That said, it is rather pathetic to paywall this) A moment's consideration ought to show that Open AI has plenty of significant work they they can be doing, even if the core model never gets any better than this.
betaminecraft 1y ago
[dead]
Havoc 1y ago
The messaging is all over the place anyway. Not so long ago OAI was talking about faster iterations and warning people to not expect huge leaps. (A position that makes sense imo). Yet people talk about AGI in a serious manner?
- coldpie 1y ago
  I don't think anyone serious is talking about AGI from LLMs, no.
  - phist_mcgee 1y ago
    And yet altman talks about AGI being imminent, but his company has only ever produced LLMs.
    - parineum 1y ago
      Now why would the CEO of an AI company say something like that!?
      - AstroBen 1y ago
        Is it a given that they need to unrealistically hype everything? To me it just seems like he's killing any and all credibility he had Probably a bad long term strategy? I mean other non-AI companies use hype too sure.. but it's maybe a little sprinkle of 1.1x on top aimed to highlight their best features. Here we're going full on 100x of reality
        parineum 1y ago
        It's not a given but Altman is a public figure for a reason while I don't know the names of any of the other CEOs off the top of my head. He talks a lot and when he talks, it's about AI. Even talking about the dangers of AI is hype because it implies it's an important topic to discuss now because it's imminent.
        coldpie 1y ago
hodgehog11 1y ago
Yes, GPT-5 is more of an iteration than anything else, and to me this says more about OpenAI than the rest of the industry. However, I think the majority of the improvements over the past year have been difficult to quantify using benchmarks. Users often talk about how certain models "feel" smarter on their particular tasks, and we won't know if the same is true for GPT-5 until people use it for a while. The "GPT-5 will show AGI" hype was always a ridiculously high bar for OpenAI, and I would argue that the quest for that elusive AGI threshold has been an unnecessary curse on machine learning and AI development in general. Who cares? Do we really want to replace humans? We should want better and more reliable tools (like Claude Code) to assist people, and maybe cover some of the stuff nobody wants to do. This desire for "AGI" is delivering less value and causing us to put focus on creative tasks that humans actually want to do, putting added stress on the job market. The one really bad sign in the launch, at least to me, was that the developers were openly admitting that they now trust GPT-5 to develop their software MORE than themselves ("more often than not, we defer to what GPT-5 says"). Why would you be proud of this?
- bluefirebrand 1y ago
  > Do we really want to replace humans? Unfortunately for a substantial number of people the answer to this question seems to be a resounding "yes"
  - gtirloni 1y ago
    With those people being business owners, investors, etc, 100% of the time. The other 99% would like automation to make their lives easier. Who wouldn't want the promised tech utopia? Unfortunately, that's not happening so it's understandable that people are more concerned than joyous about AI.
    - NoMoreNicksLeft 1y ago
      >With those people being business owners, investors, etc, 100% of the time. How can one run a business by replacing humans, if no humans are left with enough income to buy your products? I suspect that the desire to "replace humans" runs far deeper than just shortsighted business wants.
FergusArgyll 1y ago
Maybe the brain drain was real? we'll find out from gemini 3 I guess
- anonzzzies 1y ago
  For sure, but not for that reason; there is currently no one with a plan how to go from current (LLMs) to a better model. It's some 'more focused training' 'better prompting' 'agentic' 'smarter lookups' 'better tooling'. But fundamentally, this model is simply shagged out and it'll get a little better with the above, but the jump everyone is waiting for cannot happen without a new model invention.
  - FergusArgyll 1y ago
    My point is; maybe we can't prove that until deepmind gives us their best shot
    - anonzzzies 1y ago
      But aren't deepmind, in this infinite money AI times, giving it their best shot?
  - tim333 1y ago
    No one you know of but I'm sure people are thinking about it. Which reminds me that one of the most obvious failings of LLMs is they never say "I've been thinking about that and have a new idea." The thinking leaning thing needs work.
    - anonzzzies 1y ago
      I know multiple people who are working on this: there is just no progress yet that is more of the same and that won't work.
baggachipz 1y ago
> ...when they need to find any way possible to squeeze paid subscribers out of their (money losing) free user base. Also note that they're losing money on their paid subscribers.
emccue 1y ago
I asked it to make a drawing of the US with every state numbered from biggest to smallest with 1 being the largest. Maine was #89 (That is not a typo.) and Oregon was #1. OpenAI as a company simply cannot exist without a constant influx of investor money. Burning money on every request is not a viable business model. Companies built on OpenAI/Anthropic are similarly deeply unprofitable businesses. OpenAI needs to convert to a for-profit to get any more of the funding that Softbank promised (that its also unclear how Softbank itself would raise) or to get significant cash from anyone else. Microsoft can block this and probably will. It all reminds me of that Paddy's Dollars bit from it's always sunny. "We have no money and no inventory... there's still something we can do... that's still a business somehow..."
- anon191928 1y ago
  burning money worked for Uber. As long as they can IPO or get cheap debt from governm friends any valuation can work. Uber lost double digit billions as an app with no edge or anythin. It always made no sense beyond 1 billion
  - emccue 1y ago
    Uber's whole schtick is being what was an already profitable business model (Taxis) with lower overhead/easier access. That money they burned was on customer acquisition, building infrastructure, etc. The unit economics of paying to be driven to the airport or Benihanas was always net positive. They weren't losing money on every customer, even paying ones. There just isn't a business model here.
  - xnx 1y ago
    > burning money worked for Uber. TBD. Some people did well while Uber gave money away, but Uber is not net profitable over its lifetime.
    - dvfjsdhgfv 1y ago
      The way they do this in Europe is that an enterpreneur buys a fleet of cars and then gets a visa for a number of folks from Bangladesh and other areas who don not own any of these cars and ride them in turns (they also sleep like 10 in one appartment but that's a different story). The owner gets the money and distributes them to the actual drivers. Uber says they are innocent as they are not in an emplyer-employee position with any of these drivers. This model worked for the fleet owners so far because the Saudi gave enough money so that both (1) the customers were happy, (2) the cash from the ride could be divided between owners and drivers in a way these drivers complained only to a certain extent. But the last two years (the only profitable ones) are much worse, both for the drivers and fleet owners. There is still sunk cost in there, but once the cars get old enough they will need to think well whether to buy/lease the next batches.
bbstats 1y ago
Gemini 3.0 is gonna cook
- deezmofonutz 1y ago
  Yup and look at all that IP they just bought…
rvz 1y ago
Am I right to say that "AGI" was just...cancelled again? Did we just get scammed right in front of our eyes with an overhyped release and what is now an underwhelming model if the point was that GPT-5 was supposed to be trustworthy enough for serious use-cases and it can't even count or reason about letters? So much for the "AGI has been achieved internally" nonsense with the VCs and paid shills on X/Twitter bullshitting about the model before the release.
- baobun 1y ago
  How were you still eating that?
  - rvz 1y ago
    Never was. Knew it was a redefinition confidence trick years ago.
- mutkach 1y ago
  Not only "AGI" is cancelled but they also sort of admitted that so-called "scaling" "laws" don't work anymore. Scaling inference kinda still works, but obviously is bounded by context size and haystack-and-needle diminishing accuracy. So the promise of even steadily moving towards AGI is dubious at best.
VeejayRampay 1y ago
the whole event was shit, but we're all past the point where we can just say that, because the technology is now so entrenched that it's become unavoidable, so everything has now to jump through hoops to justify its existence and its greatness
ndr_ 1y ago
Some of the problems with GPT-5 in ChatGPT could actually be due to new model that is in place to route requests to the actual GPT-5 models. There are four models in the GPT-5 family, and I could reproduce the faulty "blueberry" test result only with the "gpt-5-chat" (aka "gpt-5-main") model through the API. This model is there to answer (near) instantly and it falls in the non-thinking category of LLMs. The "blueberry" test represents what they are particularly bad at (and what OpenAI set out to solve with o1). The other thinking models in the family, including gpt-5-nano, solve this correctly.
- profstasiak 1y ago
  so can we please stop talking of AGI until counting letter in a word are not hard?
kanak8278 1y ago
The most funny part of the demo was colored chats. That also behind a paywall. I was like are they become instagram
xnx 1y ago
Given the difference between GPT-3 and GPT-4, a fair numbering for "GPT-5" is probably "GPT-4.2".
Jimmc414 1y ago
I find this news very exciting to be perfectly honest. It’s finally time to build on the tech we already have.
tom_m 1y ago
We're hitting the ceiling of this algorithms already I guess. Someone will make a new and better one. That's how tech works. No worries.
starchild3001 1y ago
Latency users experience while getting their answers is a big part of the LLM experience. Well done model routing is a tremendous leap forward to minimize the latency & improve the user experience. E.g. I love Gemini 2.5 Pro. But it's darn slow (sorry GDM!). I love the latency I'm getting from 4o. The solution? Just combine them under one prompt, with well done model routing. Is GPT5 router "good enough"? We'll see. I think OpenAI is a smart company. And Sama is a tremendous leader. They're moving in the right direction.
amUsingFreeBSD 1y ago
Trust takes time to build. I think OpenAI is now finding out how fast and easy it is to lose it.

86 points by csmeyer 1y ago

The focus now is not the model, but the Product - "here we improve the usuability by removing the choice between models", "here is a better voice for tts", "here is a nice interface for previewing html" Only about 5 minutes of the whole presentation are dedicated to enterprise usage (COO in an interview sort of indirectly confirms that haven't figured it out yet). And they are cutting the costs already (opaque routing between models for non-API users is a clear sign of that). The term "AGI" is dropped, no more exponential scaling bullshit - just incremental changes over the time and only over select few domains. Actually it is a more welcoming sign and not concerning at all that this technology matures and crystallizes around this point. We will charitably forget and forgive all the insane claims made by Sam Altman in the previous years. He can also forget about cutting ties with Microsoft for that same reason.

davydm 1y ago

Issues like this are why I don't use ai agents for code. I don't want to sift through the bullshit confidently spewed out by the model. It doesn't understand anything. It can't possibly "understand my codebase". It can only predict tokens, and it can only be useful if the pattern has been seen before. Even then, it will product buggy replicas, which I've pointed out during demos. I disabled the ai helpers in my IDEs because the slop the produce is not high quality code, often wrong, often misses what I wanted to achieve, often subtly buggy. I don't have the patience to deal with that, and I don't want to waste the time on it. Time is another aspect of this conversation, with people claiming time wins, but the data not backing it up, possibly due to a number of factors intrinsic to our squishy evolved brains. If you're interested, go find gurwinder's article on social media and time - I think the same forces are at work in the ai-faithful.

mrits 1y ago
There is a threshold that every developer needs for them to make it be worth their time. For me that has already been met. Your comment makes me think that you don't believe it will start producing higher quality code than you anytime soon. I think most of us are in the camp that even though we don't need AI right now we believe we will not be valuable in the near future without being highly proficient with the tooling.
- bluefirebrand 1y ago
  > even though we don't need AI right now we believe we will not be valuable in the near future This reads to me like you don't think you're valuable right now either

cwrichardkim 1y ago

> They admitted that they were, and I am not lying about this, paywalling chat colors. […] This is a feature that a company adds when they are out of ideas This observation + sherlocking cursor suggests that perhaps sherlocking is the ideation strategy. Curious to see if they’re subsidizing token costs specifically to farm and Sherlock ideas

dentemple 1y ago
Yeah, I agree with the OP here. After all this time, being able to change the chat colors at this point has some real We-reached-the-bottom-of-the-backlog energy, and they're just now implementing the ideas that weren't considered important enough before by the PMs to consider. It hardly feels like a next generation release. As a related anecdote (not saying that this is industry standard, just pointing out my own experience), the startup I work for launched their app four years ago, and, for all four of those years, we've had "Implement a Dark Mode design" sitting at the bottom of our own backlog. Higher priority feature requests are always pre-empting it.
- msabalau 1y ago
  The core product failure here is overhyping incremental improvement, eroding trust. PMs operating at this level ought to be bringing in some low cost UX improvements alongside major features. That simply isn't a sign that they've run ought of backlog. (That said, it is rather pathetic to paywall this) A moment's consideration ought to show that Open AI has plenty of significant work they they can be doing, even if the core model never gets any better than this.

betaminecraft 1y ago

[dead]

Havoc 1y ago

The messaging is all over the place anyway. Not so long ago OAI was talking about faster iterations and warning people to not expect huge leaps. (A position that makes sense imo). Yet people talk about AGI in a serious manner?

coldpie 1y ago
I don't think anyone serious is talking about AGI from LLMs, no.
- phist_mcgee 1y ago
  And yet altman talks about AGI being imminent, but his company has only ever produced LLMs.
  - parineum 1y ago
    Now why would the CEO of an AI company say something like that!?
    - AstroBen 1y ago
      Is it a given that they need to unrealistically hype everything? To me it just seems like he's killing any and all credibility he had Probably a bad long term strategy? I mean other non-AI companies use hype too sure.. but it's maybe a little sprinkle of 1.1x on top aimed to highlight their best features. Here we're going full on 100x of reality
      - parineum 1y ago
        It's not a given but Altman is a public figure for a reason while I don't know the names of any of the other CEOs off the top of my head. He talks a lot and when he talks, it's about AI. Even talking about the dangers of AI is hype because it implies it's an important topic to discuss now because it's imminent.
      - coldpie 1y ago

hodgehog11 1y ago

Yes, GPT-5 is more of an iteration than anything else, and to me this says more about OpenAI than the rest of the industry. However, I think the majority of the improvements over the past year have been difficult to quantify using benchmarks. Users often talk about how certain models "feel" smarter on their particular tasks, and we won't know if the same is true for GPT-5 until people use it for a while. The "GPT-5 will show AGI" hype was always a ridiculously high bar for OpenAI, and I would argue that the quest for that elusive AGI threshold has been an unnecessary curse on machine learning and AI development in general. Who cares? Do we really want to replace humans? We should want better and more reliable tools (like Claude Code) to assist people, and maybe cover some of the stuff nobody wants to do. This desire for "AGI" is delivering less value and causing us to put focus on creative tasks that humans actually want to do, putting added stress on the job market. The one really bad sign in the launch, at least to me, was that the developers were openly admitting that they now trust GPT-5 to develop their software MORE than themselves ("more often than not, we defer to what GPT-5 says"). Why would you be proud of this?

bluefirebrand 1y ago
> Do we really want to replace humans? Unfortunately for a substantial number of people the answer to this question seems to be a resounding "yes"
- gtirloni 1y ago
  With those people being business owners, investors, etc, 100% of the time. The other 99% would like automation to make their lives easier. Who wouldn't want the promised tech utopia? Unfortunately, that's not happening so it's understandable that people are more concerned than joyous about AI.
  - NoMoreNicksLeft 1y ago
    >With those people being business owners, investors, etc, 100% of the time. How can one run a business by replacing humans, if no humans are left with enough income to buy your products? I suspect that the desire to "replace humans" runs far deeper than just shortsighted business wants.

FergusArgyll 1y ago

Maybe the brain drain was real? we'll find out from gemini 3 I guess

anonzzzies 1y ago
For sure, but not for that reason; there is currently no one with a plan how to go from current (LLMs) to a better model. It's some 'more focused training' 'better prompting' 'agentic' 'smarter lookups' 'better tooling'. But fundamentally, this model is simply shagged out and it'll get a little better with the above, but the jump everyone is waiting for cannot happen without a new model invention.
- FergusArgyll 1y ago
  My point is; maybe we can't prove that until deepmind gives us their best shot
  - anonzzzies 1y ago
    But aren't deepmind, in this infinite money AI times, giving it their best shot?
- tim333 1y ago
  No one you know of but I'm sure people are thinking about it. Which reminds me that one of the most obvious failings of LLMs is they never say "I've been thinking about that and have a new idea." The thinking leaning thing needs work.
  - anonzzzies 1y ago
    I know multiple people who are working on this: there is just no progress yet that is more of the same and that won't work.

baggachipz 1y ago

> ...when they need to find any way possible to squeeze paid subscribers out of their (money losing) free user base. Also note that they're losing money on their paid subscribers.

emccue 1y ago

I asked it to make a drawing of the US with every state numbered from biggest to smallest with 1 being the largest. Maine was #89 (That is not a typo.) and Oregon was #1. OpenAI as a company simply cannot exist without a constant influx of investor money. Burning money on every request is not a viable business model. Companies built on OpenAI/Anthropic are similarly deeply unprofitable businesses. OpenAI needs to convert to a for-profit to get any more of the funding that Softbank promised (that its also unclear how Softbank itself would raise) or to get significant cash from anyone else. Microsoft can block this and probably will. It all reminds me of that Paddy's Dollars bit from it's always sunny. "We have no money and no inventory... there's still something we can do... that's still a business somehow..."

anon191928 1y ago
burning money worked for Uber. As long as they can IPO or get cheap debt from governm friends any valuation can work. Uber lost double digit billions as an app with no edge or anythin. It always made no sense beyond 1 billion
- emccue 1y ago
  Uber's whole schtick is being what was an already profitable business model (Taxis) with lower overhead/easier access. That money they burned was on customer acquisition, building infrastructure, etc. The unit economics of paying to be driven to the airport or Benihanas was always net positive. They weren't losing money on every customer, even paying ones. There just isn't a business model here.
- xnx 1y ago
  > burning money worked for Uber. TBD. Some people did well while Uber gave money away, but Uber is not net profitable over its lifetime.
  - dvfjsdhgfv 1y ago
    The way they do this in Europe is that an enterpreneur buys a fleet of cars and then gets a visa for a number of folks from Bangladesh and other areas who don not own any of these cars and ride them in turns (they also sleep like 10 in one appartment but that's a different story). The owner gets the money and distributes them to the actual drivers. Uber says they are innocent as they are not in an emplyer-employee position with any of these drivers. This model worked for the fleet owners so far because the Saudi gave enough money so that both (1) the customers were happy, (2) the cash from the ride could be divided between owners and drivers in a way these drivers complained only to a certain extent. But the last two years (the only profitable ones) are much worse, both for the drivers and fleet owners. There is still sunk cost in there, but once the cars get old enough they will need to think well whether to buy/lease the next batches.

bbstats 1y ago

Gemini 3.0 is gonna cook

deezmofonutz 1y ago
Yup and look at all that IP they just bought…

rvz 1y ago

Am I right to say that "AGI" was just...cancelled again? Did we just get scammed right in front of our eyes with an overhyped release and what is now an underwhelming model if the point was that GPT-5 was supposed to be trustworthy enough for serious use-cases and it can't even count or reason about letters? So much for the "AGI has been achieved internally" nonsense with the VCs and paid shills on X/Twitter bullshitting about the model before the release.

baobun 1y ago
How were you still eating that?
- rvz 1y ago
  Never was. Knew it was a redefinition confidence trick years ago.
mutkach 1y ago
Not only "AGI" is cancelled but they also sort of admitted that so-called "scaling" "laws" don't work anymore. Scaling inference kinda still works, but obviously is bounded by context size and haystack-and-needle diminishing accuracy. So the promise of even steadily moving towards AGI is dubious at best.

VeejayRampay 1y ago

the whole event was shit, but we're all past the point where we can just say that, because the technology is now so entrenched that it's become unavoidable, so everything has now to jump through hoops to justify its existence and its greatness

ndr_ 1y ago

Some of the problems with GPT-5 in ChatGPT could actually be due to new model that is in place to route requests to the actual GPT-5 models. There are four models in the GPT-5 family, and I could reproduce the faulty "blueberry" test result only with the "gpt-5-chat" (aka "gpt-5-main") model through the API. This model is there to answer (near) instantly and it falls in the non-thinking category of LLMs. The "blueberry" test represents what they are particularly bad at (and what OpenAI set out to solve with o1). The other thinking models in the family, including gpt-5-nano, solve this correctly.

profstasiak 1y ago
so can we please stop talking of AGI until counting letter in a word are not hard?

kanak8278 1y ago

The most funny part of the demo was colored chats. That also behind a paywall. I was like are they become instagram

xnx 1y ago

Given the difference between GPT-3 and GPT-4, a fair numbering for "GPT-5" is probably "GPT-4.2".

Jimmc414 1y ago

I find this news very exciting to be perfectly honest. It’s finally time to build on the tech we already have.

tom_m 1y ago

We're hitting the ceiling of this algorithms already I guess. Someone will make a new and better one. That's how tech works. No worries.

starchild3001 1y ago

Latency users experience while getting their answers is a big part of the LLM experience. Well done model routing is a tremendous leap forward to minimize the latency & improve the user experience. E.g. I love Gemini 2.5 Pro. But it's darn slow (sorry GDM!). I love the latency I'm getting from 4o. The solution? Just combine them under one prompt, with well done model routing. Is GPT5 router "good enough"? We'll see. I think OpenAI is a smart company. And Sama is a tremendous leader. They're moving in the right direction.

amUsingFreeBSD 1y ago

Trust takes time to build. I think OpenAI is now finding out how fast and easy it is to lose it.