ChatGPT CEO Sam Altman is 'surprised' by how much people blindly trust it: 'AI hallucinates'
Are you among those who believe the AI boom to be a plague descending humanity into an irreversible descent? Or is the icon pinned on the homepage of your phone and arguably your most used app? ChatGPT Sam Altman is 'surprised' at how much people trust AI
Either way, you'll want to read this one — either for the satisfaction a good old 'I told you so!' will get you, or, because this might be the PSA you NEED. And it comes straight from the horse's mouth.
ChatGPT CEO Sam Altman spoke at the OpenAI Podcast earlier this month, and in no uncertain words, expressed genuine surprise at how much of their life people had tied into being dependent on the AI messenger.
"People have a very high degree of trust in ChatGPT, which is interesting, because AI hallucinates. It should be the tech that you don't trust that much", he said, matter-of-factly. That being said, it's been a trial and error process for the CEO himself too. Sharing a personal anecdote, Sam revealed how when he turned a new parent, he found himself a little too logistically woven into the ChatGPT nexus as he made sense of his new role: "It was always on, helping me decide everything from nap routines to what to do about diaper rash", he said.
"But I had to remind myself it doesn't always get it right", he added and that's what anybody who finds themselves on the same wavelength should ideally take note of.
As a parting shot, Sam articulated on how dependence on and in the real world, is very necessary in making sure that the power remains vested in us: "We need societal guardrails. We're at the start of something powerful, and if we're not careful, trust will outpace reliability", he said.
Running things through ChatGPT, asking it for help in maneuvering certain situations and other such instances are perfectly fine. However, it's best to take most of what comes up as suggestions and pointers instead of the final world. And just because it sounds human, doesn't in the least bit make it so. AI ethicist at the University of Toronto, Dr. Melissa Tran captures the dilemma well: "It speaks like a confident human. That alone makes people feel like it knows what it's talking about even when it doesn't".
So what's your stance on ChatGPT now?
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
22 minutes ago
- Time of India
‘We take data security seriously': Scale AI locks down sensitive files of Google, Meta and xAI that were reportedly made publicly accessible
Scale AI has secured sensitive project files of clients, including Meta and xAI , after a report suggested that thousands of them, stored in Google Docs, were publicly accessible via links. Following the report, Scale AI, which uses human gig workers to improve the AI models of major tech companies, have now secured those files. A report by Business Insider cited four Scale contractors who remained anonymous to claim that this action temporarily prevented teams of workers from opening training documents. According to the report, thousands of Scale AI files that were previously reviewed and found to be public are now private. The company also recently resolved many of the document access issues for teams and contributors who are now being granted individual access to documents, one worker confirmed. What Scale AI said about securing its clients' files from public access In a statement to BI, a Scale AI spokesperson said, 'We take data security seriously. We remain committed to robust technical and policy safeguards to protect confidential information and are always working to strengthen our practices.' The company also said that it is conducting a thorough investigation and has disabled public document sharing from its systems. However, the company has yet to comment on the specific security changes it has implemented. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like 5 Books Warren Buffett Wants You to Read In 2025 Blinkist: Warren Buffett's Reading List Undo While the report didn't mention any signs of a data breach at Scale AI, cybersecurity experts said that publicly accessible documents could expose the company to potential hacking risks. As per the previous report, Scale AI left over 85 public Google Docs accessible online, exposing sensitive and confidential information related to AI projects for clients like Google and Meta. Some documents also contained employee contact details. The issue came to light after the older report prompted Scale AI to restrict document sharing. The sudden lockdown disrupted work for many contractors, with some teams losing access mid-task. The document lockout further added to the uncertainty faced by Scale AI contractors, who were already impacted by Meta's investment and hiring of CEO Alexandr Wang for its new AI initiative. Following the deal, Google paused several projects with Scale AI, and both OpenAI and xAI also halted work with the company. Contractors reported that projects were paused without warning, and while Scale AI informed them about the Meta investment, many were reportedly not notified about the client pullbacks. iQoo Neo 10 Review: Flagship-Like Gaming Experience at Just Rs 31,999!
&w=3840&q=100)
Business Standard
an hour ago
- Business Standard
S&P 500, Nasdaq near record highs as interest rate-cut bets creep up
The S&P 500 and Nasdaq were on track to open near record levels on Thursday as robust earnings from memory-chip maker Micron fueled optimism around artificial intelligence while investors assessed a slew of economic data. The benchmark S&P 500 and tech-led Nasdaq were about 1 per cent below their all-time peaks at Wednesday's close, following a truce in the Middle East conflict. The Nasdaq 100 - a subset of the Nasdaq composite index - closed at a record high in the previous session. Semiconductor makers got a lift from a 1.3 per cent rise in Micron Technology shares premarket after the company forecast quarterly revenue above estimates, citing strong demand for its chips used in AI data-centers. Shares of Marvell Technology and Advanced Micro Devices climbed 2.3 per cent and 2.6 per cent, respectively. Nvidia rose 1.3 per cent after scaling a fresh all-time high on Wednesday. Economic data was mixed. The final reading from the U.S. Commerce Department showed gross domestic product contracted 0.5 per cent in the first quarter. Economists polled by Reuters had forecast a 0.2 per cent contraction. "A 0.5 per cent contraction is worse-than-expected, but that really reflects how much imports overwhelmed exports in the first quarter, as businesses and consumers tried to get ahead of the tariff program," said Sam Stovall, chief investment strategist at CFRA Research. At 08:52 a.m. ET, Dow E-minis were up 100 points, or 0.23 per cent, S&P 500 E-minis were up 18.5 points, or 0.3 per cent. Nasdaq 100 E-minis were up 90 points, or 0.4 per cent Investors also weighed Federal Reserve Chair Jerome Powell's remarks during his two-day congressional testimony, where he reiterated a wait-and-see approach toward interest rate cuts. President Donald Trump intensified his attacks on Powell for not cutting rates sooner. On Wednesday, Trump said he had narrowed down his choice to three or four candidates to potentially replace the Fed chair. Traders are pricing in about 63 basis points worth of rate cuts by the end of 2025, with expectations for the first cut in September. On Friday, the Personal Consumption Expenditures report - the Fed's preferred gauge of inflation - will be scrutinized to ascertain tariff-induced price changes in the U.S. economy. Some of the central bank officials scheduled to speak later in the day include Fed Cleveland President Beth Hammack, Fed Board Governor Michael Barr and Fed Minneapolis President Neel Kashkari. Shares of sportswear company Nike edged up 0.6 per cent ahead of its quarterly results.
&w=3840&q=100)
First Post
an hour ago
- First Post
As race for AI heats up, Meta poaches three OpenAI researchers
Sam Altman recently said Mark Zuckerberg has been personally reaching out to top AI scientists, messaging them via WhatsApp and organising meetings read more Meta CEO Mark Zuckerberg presents Meta AI with Voice, as he makes a keynote speech during the Meta Connect annual event, at the company's headquarters in Menlo Park, California, US, on September 25, 2024. (Photo: Reuters) Meta has poached three senior researchers from OpenAI, marking a notable gain in its ongoing effort to attract top artificial intelligence talent, even as OpenAI CEO Sam Altman has publicly dismissed the tactics used by Meta chief executive Mark Zuckerberg. The three researchers, Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai, previously established OpenAI's Zurich office and have now joined Meta's recently formed superintelligence team, The Wall Street Journal recently reported. Their hiring comes amid a high-profile recruiting campaign by Zuckerberg, who, according to reports, has been offering compensation packages worth more than $100 million in attempts to lure AI researchers away from OpenAI and other competitors. STORY CONTINUES BELOW THIS AD Sam Altman mocks Mark Zuckerberg's tactics Altman, in a recent podcast with his brother Jack, said Zuckerberg has been personally reaching out to top AI scientists, messaging them via WhatsApp and organising meetings through a chat group reportedly dubbed 'Recruiting Party'. The Journal reported that Zuckerberg has hosted prospective hires at his homes in Palo Alto and Lake Tahoe. The aggressive approach has had mixed results. Meta recently secured a $14 billion partnership with Scale AI, bringing CEO Alexandr Wang on board. But efforts to recruit OpenAI co-founders Ilya Sutskever and John Schulman were unsuccessful. Both have launched their own startups in recent months. In the same podcast, Altman said, 'I'm really happy that, at least so far, none of our best people have decided to take him up on [those offers].' Meta's big spending plans Meta plans on splurging up to a whopping $65 billion on capital expenditures this year. The amount would be to largely to build out its AI ambitions. Over the course of last couple of months, Zuckerberg has tried to project the vision of a future where people speak to friends that are AI, use AI to create advertising campaigns from scratch, and get in touch with AI business agents when they need to speak to a brand.
