Microsoft AI chief, Mustafa Suleyman, says firms should stop calling AI conscious
He also said that these cases aren't just limited to people who are already at risk of mental health issues.
'My central worry is that many people will start to believe in the illusion of AIs as conscious entities so strongly that they'll soon advocate for AI rights, model welfare and even AI citizenship. This development will be a dangerous turn in AI progress and deserves our immediate attention,' he noted.
Suleyman called the illusion that AI systems were feeling, thinking beings, Seemingly Conscious AI just because they mimicked human responses well.
He also added that the industry needs to collaborate on the best practices for implementing guardrails to prevent this perception of consciousness or 'refute that perception if a user develops it.'
AI companions, which are an entirely new category needed these guardrails urgently, he said.
Although AI firms like OpenAI have repeated that these extreme cases of psychosis caused by AI were rare and a small fraction of their userbase, there have been multiple incidents that have drawn attention to the behaviour of AI chatbots.
Post the launch of GPT-5 on August 7, there was a huge backlash after users complained about GPT-4o being deprecated abruptly. A considerable fraction of these protesting users had formed emotional attachments to the AI model.
OpenAI was forced to bring the AI model back after the reaction.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
&w=3840&q=100)
Business Standard
an hour ago
- Business Standard
Black box to sandbox: Framework needed for safe AI adoption in finance
The Monetary Authority of Singapore (MAS) introduced "FEAT" (fairness, ethics, accountability, and transparency) principles in 2018 Business Standard Editorial Comment Mumbai Listen to This Article The Reserve Bank of India's (RBI's) Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE-AI) committee released its final report last week, proposing guidelines for the responsible use of artificial intelligence (AI) in the financial sector. The framework is anchored around seven guiding 'sutras' or principles — trust, people first, innovation, fairness, accountability, explainability, and resilience — and is backed by 26 actionable recommendations across six pillars — infrastructure, policy, capacity, governance, protection, and assurance. The report arrives at a critical juncture. A World Economic Forum study (2025) estimates that global AI investment in finance could reach $97 billion
&w=3840&q=100)
Business Standard
2 hours ago
- Business Standard
US examines equity stake in chip makers for CHIPS Act cash grants: Report
US Commerce Secretary Howard Lutnick is looking into the government taking equity stakes in Intel and other chipmakers in exchange for grants under the CHIPS Act, which aims to spur factory-building in the US, two sources said. As part of a plan to revive US manufacturing - a key Trump agenda - Lutnick said earlier on Tuesday the US government wants an equity stake in Intel in exchange for cash grants approved by the administration of former President Joe Biden. Now Lutnick wants to expand that plan to other companies, according to a White House official and a person familiar with the situation. The Trump administration has recently made unusual deals with US companies, including allowing AI chip giant Nvidia to sell its H20 chips to China in exchange for the US government receiving 15 per cent of those sales. The Pentagon is slated to become the largest shareholder in a small mining company to boost output of rare earth magnets. The government's intervention in corporate matters has worried critics who say President Donald Trump's actions create new categories of corporate risk and that a bad bet could mean a hit to taxpayer funds. Much of the funding under the CHIPS Act has not yet been dispersed for companies such as Micron, Taiwan Semiconductor Manufacturing Co, Samsung and Intel. TSMC and Intel declined to comment. Micron, Samsung and the White House did not respond to requests for comment on whether Lutnick is considering more stakes. The two sources told Reuters on Tuesday that Treasury Secretary Scott Bessent is also involved in the CHIPS Act discussions, but that Lutnick is driving the process. The Commerce Department oversees the $52.7 billion CHIPS Act money. Lutnick has been pushing the equity idea, the sources said, adding that Trump likes the idea. White House Press Secretary Karoline Leavitt confirmed earlier that Lutnick was working on a deal with Intel to take a 10 per cent government stake. "The president wants to put America's needs first, both from a national security and economic perspective, and it's a creative idea that has never been done before," she told reporters. Speaking on CNBC, Lutnick said the US wants a return on its "investment". "We'll get equity in return for that ... instead of just giving grants away," he said. Trump has previously said he wanted to kill the CHIPS Act program. Lutnick's comments suggested any stake would be non-voting, meaning it would not enable the US government to tell the company how to run its business. His comments came a day after SoftBank Group agreed to invest $2 billion in Intel, which has struggled to compete after years of management blunders. "The Biden administration literally was giving Intel money for free and giving TSMC money for free, and all these companies just giving the money for free, and Donald Trump turned it into saying, 'Hey, we want equity for the money. If we're going to give you the money, we want a piece of the action for the American taxpayer'," Lutnick said. South Korean presidential advisor Kim Yong-beom said neither the government nor the potentially affected companies have heard about such a plan. He added that foreign companies like Samsung needed "predictability" for their US investments. A Korean chip industry official, meanwhile, said it would be hard for chipmakers to accept US government equity stakes, and some may either decide not to invest or delay investments unless Washington provides incentives like increasing funding. Taking lawmaker questions in Taipei on Wednesday and asked whether the US government could take a stake in TSMC, Taiwan Economy Minister Kuo Jyh-huei said his ministry would consult with the company, which he pointed out was private and not a state-owned enterprise. "We will also discuss with the National Development Council, as it is a shareholder of TSMC. We will thoroughly understand the underlying meaning of the US Commerce Secretary's remarks, but this will require some time for discussion and assessment," Kuo said.
Economic Times
2 hours ago
- Economic Times
Cracks in the AI hype? Warning signs suggest the Bubble could soon explode
In March 2000, Barron's asked, 'When will the internet bubble burst?' warning the pop could happen before year-end. On that same day, a top tech company saw its share price drop 60pc, and many others collapsed, wiping out trillions. Now, some Wall Street experts worry the same 'unpleasant popping sound' may hit the AI boom. On Tuesday, tech stocks fell sharply after a report from MIT said most AI investments give 'zero return' for businesses. MIT academics wrote: 'Despite $30-40bn in enterprise investment into Generative AI, 95pc of organisations are getting zero return'. Nvidia shares dropped 3.5pc and Palantir fell 9pc after the report. MIT's findings could be the pin that pops the tech stock bubble, which has increased US stock values by trillions. Since ChatGPT's launch in 2022, Silicon Valley has claimed AI chatbots will transform the economy, promising cost savings and productivity gains. MIT's report suggests the AI revolution has stalled, despite big investments, as reported by The Telegraph. MIT surveyed 150 business leaders and 350 employees, finding only 5pc of AI pilots generate millions in value; the rest show no measurable profit impact. Marko Kolanovic, former JP Morgan head of research, commented: 'Sounds about right for a bubble'. MIT found half of AI projects fail, 80pc of companies explored AI, but only 40pc actually deployed it. Only 20pc of enterprise-grade AI systems reached the pilot stage and just 5pc reached production. ALSO READ: Google Pixel 10 series and Pixel 10 Pro fold launched with AI features and new designs Many employees prefer using consumer AI products like ChatGPT on their own rather than corporate AI tools. The report emphasized: 'AI is already transforming work, just not through official channels'. Despite benefiting from AI, OpenAI CEO Sam Altman warned investors: 'Are investors overexcited? My opinion is yes… some people stand to lose a phenomenal amount of money'. Investors worry about SoftBank, which invested billions in OpenAI; its shares fell 7pc on Tuesday. OpenAI recently launched ChatGPT-5, which underwhelmed users; improvements were seen as small or incremental. Morgan Stanley predicts $3tn will be spent on data centres over three years, mostly fueled by debt, to support AI growth, as stated by The Telegraph. Morgan Stanley also projects AI could add $16tn to the S&P 500 through 40pc salary savings, though MIT's report suggests this may be unrealistic. Meta announced a reorganization of its AI division, including staff reductions, showing even big AI believers are cautious. Mark Zuckerberg has spent hundreds of millions to hire AI engineers at Meta, as per the New York Times the sell-off, the market hasn't collapsed completely; some analysts see it as just a speed bump. Dan Ives, tech analyst at Wedbush Securities, said skeptics of the tech rally will be proven wrong again. Nvidia will report results next week, potentially showing the true state of AI investments in major companies. Ives believes the tech bull cycle will continue for another 2-3 years, but many are still listening for the ominous pop, as mentioned by The Telegraph. Q1. Is the AI boom at risk of a market crash? Yes, experts warn that most AI investments show zero return, which could trigger a tech stock bubble burst. Q2. Are big companies slowing down their AI projects? Yes, many firms like Meta and SoftBank are cautious, with AI projects failing or being scaled back.
