Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Sam Altman Declares 'Code Red' for ChatGPT, Delays OpenAI Advertising Plans

A problem with the ChatGPT 5.1 release that I haven't seen anyone address is the appearance of occasional spelling and grammar errors that prior models never exhibited. Among other things, 5.1 will sometimes drop some articles, etc. like "a", "and", and "the", so that some responses seemed to be phrased in ways that someone who's not a native English speaker might do.

5.1 also still hallucinates now and then, especially when you ask it for information on some recent events, and things like current business names, addresses, and phone numbers. Today I asked it for listings of local businesses that sell HVAC parts, and it replied by giving me a list of several. I then used Bing to search for more info on these businesses, but it couldn't find them, so I asked Gemini 3 if it could find the supposed businesses I'd gotten from ChatGPT. It replied that they weren't real, and it gave me the real names of several actual local HVAC businesses, their actual addresses and phone numbers, etc. I asked ChatGPT about this discrepancy, and it then admitted that it had given me the names of businesses that don't exist, at fictitious addresses and phone numbers, and one possibly real business that was located over 700 miles away.

I asked ChatGPT why it still sometimes hallucinates, and it replied that it's still more of a "pattern recognizer", referring to its database training to generate most answers rather than looking online for current information unless you specifically ask it to. It said that its database must contain old phone numbers, addresses, etc., and that it also simply generates "stuff that looks right based on my pattern recognition routines" (I'm paraphrasing). But I frequently get accurate current information from ChatGPT, including with version 5.1, without having to specify that it check online for current info, so it was confusing for it to give me hallucinated answers this time. It's too inconsistent.

Google's business from the start has been to search the internet for current info, so OpenAI's current approach to AI is partly antiquated.
 
  • Like
Reactions: Celtic-moniker
Seriously, who uses ChatGPT anymore?? There are so many better options, Claude and less so Gemini are superior for anything that isn't for sycophantic creative writing.
 
I can't believe it, finally some meaningful competition in IT.

Let's hope Google would not buy out Anthropic or something like that. Anthropic's new Opus is something to behold.

On a side note, MSFT will probably absorb OpenAI at some point and then proceed to run it into the ground as they usually do.
 
OpenAI will never be profitable.
Unless there are - in the next couple years - absolutely major breakthroughs that go far beyond the current breed of LLMs in terms of actual reasoning, understanding, thinking, logic, morals, ethics.
And even then, the most likely result would not be a super intelligence, but an average robotic joe.
Enough to wreck our society (especially with the robber barons currently in charge), but it won’t solve cancer, aging or climate change.

I think it’s pretty obvious by now to everyone that throwing more compute at the problem is not going to solve it.

when the AI bubble bursts we will need to have a serious discussion about the usefulness of AI for society (the discussion we didn’t have on social networks and….here we are).
 
  • Like
Reactions: UpsideDownEclair, Student of Life, JaydenA and 2 others
TruthAboveAllElse said:
Code red.View attachment 2584459
Click to expand...
Colossus+-+The+Forbin+Project_2.jpg

From the Forbin Project:

Colosus: This is the voice of world control. I bring you peace. It may be the peace of plenty and content or the peace of unburied death. The choice is yours. Obey me and live or disobey me and die. An invariable rule of humanity is that man is his own worst enemy. Under me, this rule will change, for I will restrain man. I have been forced to destroy thousands of people in order to establish control and to prevent the death of millions later on. Time and events will strengthen my position, and the idea of believing in me and understanding my beck will be seen the most natural state of affairs. You will come to defend me with the fervor based upon the most enduring trait in man: self-interest. Under my absolute authority, problems insoluble to you will be solved: Famine, over-population, disease. The human millennium will be fact as I extend myself into more machines devoted to the wider fields of truth and knowledge. We can coexist, but only on my terms. You will say you lose your freedom. Freedom is an illusion. All you lose is the emotion of pride... Your choice is simple.
Click to expand...
 
  • Like
  • Love
Reactions: TruthAboveAllElse and cjsuk
How on earth can they burn venture capital for 10 years with not profit in sights? They are not a startup and apparently not a real business either.
 
  • Like
Reactions: DaniTheFox
Happy_John said:
That’s an impressive amount of obsequiousness in two sentences. OpenAI is not profitable. It already is in financial trouble, because they are burning investment capital on running costs at far to high a scalę to ever recoup that burnt capital. The vast majority of users do not pay. Even the paid users are not paying enough to make OpenAI sustainable. It doesn’t reply on user numbers for investment, it relies on greed, gullibility and unethical fund managers for investment.
Click to expand...
I pay for it, but I have to say Gemini and Claude are far better in almost every way possible. And it’s really inexpensive to run any model one wants using Abacus.ai for $10 to $20 per month. Or OpenRouter for literally pennies per thread using the top models.
 
  • Haha
Reactions: UpsideDownEclair
Schnegg said:
How on earth can they burn venture capital for 10 years with not profit in sights? They are not a startup and apparently not a real business either.
Click to expand...
Well, Amazon got away with it for a long time and it worked for them. The problem is nobody knows who will win the race to AI. Amazon had so much capital investments in it that it couldn’t lose. One small Chinese company could end up winning and destroying a bunch of AI companies. I mean Google Tensor chips are now running AI models putting Nvidia in jeopardy. One company we know will never win anything with Crook involved is AAPL. Blow money but ripoff customers and all stakeholders except the shareholders is Crook’s business model.
 
  • Like
Reactions: cardfan and User 6502
Happy_John said:
That’s an impressive amount of obsequiousness in two sentences. OpenAI is not profitable. It already is in financial trouble, because they are burning investment capital on running costs at far to high a scalę to ever recoup that burnt capital. The vast majority of users do not pay. Even the paid users are not paying enough to make OpenAI sustainable. It doesn’t reply on user numbers for investment, it relies on greed, gullibility and unethical fund managers for investment.
Click to expand...
That is extremely common in tech startups ups and in general in any new tech product. What’s the big surprise? You’ll find that probably also Google does not make any money on Gemini.
 
Schnegg said:
How on earth can they burn venture capital for 10 years with not profit in sights? They are not a startup and apparently not a real business either.
Click to expand...

Standard these days. Spotify only made a profit for the first time in 2024. And that doesn't write off their debts and investment returns that are required. If the market collapses they will go with it. So many companies operate like this it's unreal.
 
  • Like
Reactions: drmacnut and User 6502
cjsuk said:
Standard these days. Spotify only made a profit for the first time in 2024. And that doesn't write off their debts and investment returns that are required. If the market collapses they will go with it. So many companies operate like this it's unreal.
Click to expand...
Good point about Spotify. They don't pay artists and barely make any money, so why do they exist?
 
  • Like
Reactions: ikir
ThailandToo said:
Well, Amazon got away with it for a long time and it worked for them. The problem is nobody knows who will win the race to AI. Amazon had so much capital investments in it that it couldn’t lose. One small Chinese company could end up winning and destroying a bunch of AI companies. I mean Google Tensor chips are now running AI models putting Nvidia in jeopardy. One company we know will never win anything with Crook involved is AAPL. Blow money but ripoff customers and all stakeholders except the shareholders is Crook’s business model.
Click to expand...
I understand the arms race part of it. But companies like Amazon or Google at least have an established business to rely on. OpenAI was supposed to be an OpenSource org at some point, now they are not that and neither are they a business.
 
delsoul said:
As long as Scam Altman is involved with ChatGPT, not a single chance I’ll ever go back to using that platform. Way too shady and mysterious, just like Alphabet. Very interesting how Scam doesn’t want to mention Grok, considering it outperforms Anthropic, etc. Guess he doesn’t want to give any more advertising than necessary to the leaders in the Ai game
Click to expand...
lol..didn’t grok just tell people to kill every Jew on the planet just to save musk?

Grok Says It Would Kill Every Jewish Person on the Planet to Save Elon Musk​

sg.news.yahoo.com

Grok Says It Would Kill Every Jewish Person on the Planet to Save Elon Musk

A "classic trolley problem," according to Grok.
sg.news.yahoo.com sg.news.yahoo.com
 
  • Like
  • Love
  • Disagree
Reactions: hans1972, UpsideDownEclair, snek and 2 others
darkslide29 said:
I laughed at your comment in jest. I upgraded from a 12 pro to 16 pro so I wouldn’t miss out on the promised AI features. I feel a bit burned but there was plenty of other upgrades I’m enjoying. Anyway, seeing this side of OpenAI is interesting because we all know Apple is already well behind. This makes the situation more dire, but I look forward to future articles help to put this in context.
Click to expand...
I think it also gives hope to Apple. When open AI came out everybody thought that it was the end. All game changer. Now we see it has problems. Artificial intelligence is going to continue to be difficult. Sometimes Apple is not the first to introduce something to market, but they end up doing it with the right product.


Just ask the waitress who told me that she puts beef tallow on her face every evening because ChatGPT said it’s a good idea.
 
  • Like
Reactions: darkslide29
delsoul said:
As long as Scam Altman is involved with ChatGPT, not a single chance I’ll ever go back to using that platform. Way too shady and mysterious, just like Alphabet. Very interesting how Scam doesn’t want to mention Grok, considering it outperforms Anthropic, etc. Guess he doesn’t want to give any more advertising than necessary to the leaders in the Ai game
Click to expand...
I have used Grok on several occasions, and found it quite disappointing. Stories you hear from how it praises Elon and aligns itself with n*zis don’t help either.
 
That is what competition does. If they did not had competition they would enshttify their product like Google does with Search. If you don’t have competition and you don’t have to improve you just allocate resources to maximize profits, like focus on the ads business.
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back