OpenAI releases new ChatGPT version with 'PhD level' expertise, but it can't spell 'blueberry'
ChatGPT-5 is rolling out free to all users of the AI tool, which is used by nearly 700 million people weekly, OpenAI said in a briefing with journalists.
Co-founder and chief executive Sam Altman touted this latest iteration as 'clearly a model that is generally intelligent.'
Altman cautioned that there is still work to be done to achieve the kind of artificial general intelligence (AGI) that thinks the way people do.
'This is not a model that continuously learns as it is deployed from new things it finds, which is something that, to me, feels like it should be part of an AGI,' Altman said.
'But the level of capability here is a huge improvement.'
The rollout has not been without issues though.
ChatGPT-5 has struggled to give correct answers to simple prompts since it was released.
It even denies its own existence when asked about it.
In response to the question, 'Is this ChatGPT-5?', the large language model replied:
'You're currently chatting with ChatGPT-4o architecture, which is part of the GPT-4 family – not ChatGPT-5.'
A screenshot of ChatGPT-5 telling the user it is not ChatGPT-5.
Screenshot taken by The Journal
Screenshot taken by The Journal
Another example of the new version of the chatbot making a basic error has prompted derision from some social media users.
ChatGPT-5 cannot spell the word 'blueberry'.
When users asked how many times the letter 'b' appears in the word, it replied that there are three.
I had to try the 'blueberry' thing myself with GPT5. I merely report the results.
[image or embed]
— Kieran Healy (
@kjhealy.co
)
August 8, 2025 at 1:04 AM
Industry analysts have heralded the arrival of an AI era in which genius computers transform how humans work and play.
'As the pace of AI progress accelerates, developing superintelligence is coming into sight,' Meta chief executive Mark Zuckerberg wrote in a recent memo.
Advertisement
'I believe this will be the beginning of a new era for humanity.'
Altman said there were 'orders of magnitude more gains' to come on the path toward AGI.
'Obviously… you have to invest in compute (power) at an eye watering rate to get that, but we intend to keep doing it.'
Tech industry rivals Amazon, Google, Meta, Microsoft and Elon Musk's xAI have been pouring billions of dollars into artificial intelligence since the blockbuster launch of the first version of ChatGPT in late 2022.
Chinese startup DeepSeek shook up the AI sector early this year with a model that delivers high performance using less costly chips.
'PhD-level expert'
With fierce competition around the world over the technology, Altman said ChatGPT-5 led the pack in coding, writing, health care and much more.
'GPT-3 felt to me like talking to a high school student — ask a question, maybe you get a right answer, maybe you'll get something crazy,' Altman said.
'GPT-4 felt like you're talking to a college student; GPT-5 is the first time that it really feels like talking to a PhD-level expert in any topic.'
Altman expects the ability to create software programs on demand — so-called 'vibe-coding' – to be a 'defining part of the new ChatGPT-5 era.'
In a blog post, British AI expert Simon Willison wrote about getting early access to ChatGPT-5.
'My verdict: it's just good at stuff,' Willison wrote.
'It doesn't feel like a dramatic leap ahead from other (large language models) but it exudes competence – it rarely messes up, and frequently impresses me.'
However, Musk wrote on X, formerly Twitter, that his Grok 4 Heavy AI model 'was smarter' than ChatGPT-5.
Honest AI?
ChatGPT-5 was trained to be trustworthy and stick to providing answers as helpful as possible without aiding seemingly harmful missions, according to OpenAI safety research lead Alex Beutel.
'We built evaluations to measure the prevalence of deception and trained the model to be honest,' Beutel said.
ChatGPT-5 is trained to generate 'safe completions,' sticking to high-level information that can't be used to cause harm, according to Beutel.
The company this week also released two new AI models that can be downloaded for free and altered by users, to challenge similar offerings by rivals.
The release of 'open-weight language models' comes as OpenAI is under pressure to share inner workings of its software in the spirit of its origin as a nonprofit.
With reporting from David Mac Redmond
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Journal
an hour ago
- The Journal
New Garda aircraft kitted out with high-tech police surveillance equipment in Austria
THE GARDA'S NEW surveillance plane has been fitted out with high-tech kit in Austria in the last week, The Journal has learned. Delivery of the new de Havilland Canada-6 Twin Otter Guardian 400 aircraft is anticipated in the coming months. The aircraft is to be in Ireland before the country's six-month EU presidency from July 2026 and is set to replace the current plane, the Britton Norman Defender which has reached the end of its service life. It is understood the cost of the project will be in the region of €7 million, which will include the aircraft and the surveillance technology. It is understood that the plane has been in Austria for a number of weeks as cameras and other surveillance equipment were attached. It has now returned to Switzerland where it will be painted before making its journey to Ireland. The Austrian company is understood to be fitting the tech and internal work spaces for garda members trained to monitor the systems. Sources have said this includes high-end police role equipment such as phone tracking technology, cameras and other surveillance technology such as capabilities to monitor shipping traffic. Advertisement It is likely that the aircraft can stay in the air above an operation for a lengthy time, much more than the helicopter – known as loiter time, this capability is between five and seven. An Garda Síochána owns and operates the aircraft which will be based in the Irish Air Corps base at Casement Aerodrome in South County Dublin. It will be staffed by gardaí as observers but it will be flown by Air Corps pilots – the maintenance is done by a civilian contractor. A tendering competition is also underway for a civilian contractor to carry out maintenance of the aircraft at Casement Aerodrome and at other unspecified locations. The mission for the aircraft will be at a variety of incidents including surveillance of moving targets such as cars. Garda tender documents state that the primary mission of GASU is to provide a patrol and response capability. The method of working is known as 'visual contact flights'. The aircraft will be required to work by day and by night and will be used to 'identify a target, establish and maintain visual contact with the target'. A statement has been requested from An Garda Síochána. Readers like you are keeping these stories free for everyone... A mix of advertising and supporting contributions helps keep paywalls away from valuable information like this article. Over 5,000 readers like you have already stepped up and support us with a monthly payment or a once-off donation. Learn More Support The Journal
Irish Times
8 hours ago
- Irish Times
Is OpenAI really worth $500bn?
ChatGPT maker OpenAI is reportedly exploring a private share sale that could value the company at $500 billion (€430 billion). Is it worth it? The bullish argument: ChatGPT already has 700 million weekly users. OpenAI is pushing into search, cloud, software and hardware. It could become the foundational platform for the AI era, like Windows was for the PC. 'This is an AI revolution, and OpenAI is the golden child,' says high-profile Wedbush analyst and AI enthusiast Dan Ives. READ MORE If you are worried about valuations, says Ives, you'd have missed every technology rally in history. The bearish argument: you can't ignore valuation, and OpenAI's has shot up from $300 billion to $500 billion, valuing it at about 25 times sales. That's rich – very, very rich. Perhaps, but frothy valuations often signal that enthusiasm, not fundamentals, is calling the tune. Fund giant GMO notes the S&P 500 is up 27 per cent since sentiment flipped in early April, but more speculative corners have gone vertical. A basket of loss-making tech stocks is up 57 per cent; AI stocks, 60 per cent; highly-shorted stocks, 68 per cent; meme stocks, 77 per cent; and bitcoin-sensitive names, 112 per cent. Animal spirits are back, and momentum is driving prices more than profits. In that context, OpenAI's valuation isn't a bizarre aberration. Rather, it's a reflection of a wider market mood, with excitement trumping caution.
Irish Times
19 hours ago
- Irish Times
Thousands of ChatGPT conversations accessible online despite efforts to erase them
More than 100,000 conversations between users of ChatGPT and the online AI platform remain accessible through internet searches despite attempts by ChatGPT developer OpenAI to erase them after what was described as a 'short-lived experiment', it has been reported. Online investigative journalism unit Digital Digging said the majority of the conversations, which are preserved in their entirety, are 'harmless', but some were not. In its report, it cites the examples of an Italian lawyer discussing how to obtain land from an indigenous community in the Amazon for a multinational seeking to build a hydroelectric facility, an Egyptian seeking to use AI to write critically about the country's authoritarian regime and a researcher documenting their academic fraud. The report follows an incident at the end of last month when it became apparent that thousands of ChatGPT users were essentially ticking a box that made their conversations with the platform visible to search engines like Google without users appreciating the consequences. READ MORE The facility, which was described by a senior OpenAI executive as a 'short-lived experiment', was quickly disabled and some 50,000 conversations 'scrubbed' by the company. Experts say the Digital Digger article highlights the difficulty with erasing online posts once they have initially been made publicly available. The wider incident points to issues with AI literacy among the rapidly growing number of people engaging with the technology. The consequences can go well beyond ridicule or embarrassment, according to Pradeep Sanyal, a San Francisco-based adviser to company boards and chief executives on AI strategies. Influencers beware, Revenue is on the prowl Listen | 48:38 'Some posts reveal commercially sensitive information, legal strategies, political dissent in authoritarian contexts and confidential personal matters such as health conditions,' he said. 'These could lead to reputational damage, competitive disadvantage, regulatory scrutiny or even personal safety risks depending on jurisdiction. 'There have been revelations that, if tied back to the individual or organisation, could have significant consequences. The example of a lawyer discussing strategy to displace indigenous communities for the lowest possible price, for instance, is not merely embarrassing; it could have legal and ethical implications,' he told The Irish Times. Barry Scannell, an AI law and policy partner at William Fry solicitors and a member of the Government-appointed AI Advisory Council, said his 'jaw hit the floor' when he saw some of the original material made accessible to routine Google searches. On Monday he said that the enduring issues raised by the incident 'have shone a light on the nature of information people are sharing with AI tools, some of it deeply commercially sensitive where they look to have it help them strategise or deal with other business issues [and] some of it deeply personal, where they are treating AI like a therapist. 'Some of this reminds me of the early days of electronic communication when some people were very flippant about what they put in emails and that was sometimes shown up in discovery processes. It is certainly another reminder of the importance of companies having very clear processes and policies when it comes to their people using this technology.'
