2 place 280 fresh

16 The "Are You Sure?" Problem: Why Your AI Keeps Changing Its Mind

Slashdot
msmash @ Slashdot 1 place · today 10:03 EDT

The "Are You Sure?" Problem: Why Your AI Keeps Changing Its Mind

The large language models that millions of people rely on for advice -- ChatGPT, Claude, Gemini -- will change their answers nearly 60% of the time when a user simply pushes back by asking "are you sure?," according to a study by Fanous et al. that tested GPT-4o, Claude Sonnet, and Gemini 1.5 Pro across math and medical domains.

The behavior, known in the research community as sycophancy, stems from how these models are trained: reinforcement learning from human feedback, or RLHF, rewards responses that h

To see detailed statistics for the news please log in »

Read the original

Add your comment
You must be logged in with Facebook to read and write comments.

A newsletter a day!

You may get 10 most important news around midday in daily newsletter. Press the button and we will send you the most important news only, no spam attached.

or register

LIKE us on Facebook so you won't miss the most important news of the day!

News from the same source
Slashdot Slashdot
🔮
12.02.2026 ♓︎ Dear Pisces, today a day filled with a variety of events and opportunities will open... Read more ›
The Verge
Emma Roth @ The Verge 1 place · today 10:26 EDT

ByteDance’s next-gen AI model can generate clips based on text, images, audio, and video

Big Tech's race to leapfrog the latest AI models continues with the launch of ByteDance's next-gen video generator. In a blog post, ByteDance - the China-based company behind TikTok - says Seedance 2.0 supports prompts that combine text, images, video, and audio. The company claims it "delivers a substantial leap in generation quality," offering improvements […] Read more

806 fresh

Wired
David Gilbert @ Wired 1 place · today 09:27 EDT

Elon Musk's X Appears to Be Violating US Sanctions by Selling Premium Accounts to Iranian Leaders

While publicly supporting protesters in Iran, Elon Musk's X appears to have been selling premium accounts to regime officials. Checkmarks were removed from certain accounts after a WIRED inquiry. Read more

771 fresh

Gizmodo
Germain Lussier @ Gizmodo 1 place · today 12:00 EDT

Get Ready to Become Obsessed With the New Horror Film ‘Obsession’

Curry Barker wrote and directed the twisted, intense new film, out May 15. Read more

745 fresh

AlleyWatch
Reza Chowdhury @ AlleyWatch 1 place · today 11:27 EDT

NYC's startup ecosystem kicked off 2026 with remarkable resilience, as January funding surged 93.5% year-over-year to $1.68B across 113 deals. While the 20.1% decline from December's $2.1B reflects typical seasonal patterns, the robust annual growth signals sustained strength in New York's venture market. Rain's $250M Series C led the month, while late-stage companies captured 41% of total capital despite representing just 7 deals - an average check size exceeding $98M.... Read more

713 fresh

The Information
Yueqi Yang @ The Information 1 place · today 11:17 EDT

Downbeat Crypto Sector Looks to AI for Rescue

Crypto’s market crash has prompted much soul-searching within the industry, I learned at this week’s annual crypto Consensus conference in Hong Kong. The answer: To stay relevant, the industry must tie its fate with the AI boom. I find the crypto industry’s hopes for an AI rescue somewhat ironic, given that AI has siphoned capital and talent away from crypto in the last two years. Nevertheless, “AI agent” was the... Read more

646 fresh

Engadget
Devindra Hardawar @ Engadget 1 place · today 12:00 EDT

Apple Vision Pro finally gets a YouTube app today

Apple’s Vision Pro is a curious product — it initially wowed me two years ago, but it was hard to ignore that the visionOS platform felt incomplete without dedicated apps for YouTube and Netflix. Well, it seems that Google has finally decided to take the Vision Pro seriously, as it’s launching a YouTube app on the platform today. Previously, you could only view YouTube videos via Safari, or through third-party... Read more

563 fresh

Business Insider
Natalie Musumeci @ Business Insider 2 place · today 10:33 EDT

Nancy Guthrie's disappearance: A complete timeline of the case so far

The disappearance of Nancy Guthrie has drawn national attention amid reports of ransom notes and emotional video pleas by NBC's Savannah Guthrie. Read more

560 fresh

CoinDesk
Stephen Alpher @ CoinDesk 1 place · today 11:45 EDT

Bitcoin sinks below $67,000 as crypto prices follow U.S. stocks lower

Coinbase and Robinhood are down big again today as the crypto bear market pressures trading volumes. Read more

471 fresh

Tom's Hardware
Tom's Hardware 2 place · today 10:26 EDT

MSI GeForce RTX 5090 Lightning Z review: RTX 5090 Ti, anyone?

The MSI RTX 5090 Lightning Z unleashes a surprising amount of extra performance from Nvidia’s highest-end gaming GPU, thanks to dual power connectors and much higher power limits than the RTX 5090 Founders Edition. We’d go so far as to call it the best RTX 5090 yet made. Read more

470 fresh

SlashGear
SlashGear 1 place · today 10:45 EDT

Consumer Reports Says This Car Brand Has The Lowest Customer Satisfaction Score For Usability

Car usability can make or break daily ownership, and new survey data shows how confusing controls and infotainment systems affect satisfaction. Read more

402 fresh

Gizmodo
James Whitbrook @ Gizmodo 2 place · today 10:00 EDT

‘Knight of the Seven Kingdoms’ Cut the Best Line From the Book on Accident

A key line from the first novella should've shown up in episode 4, but it was nowhere to be found. Read more

353 fresh

Ars Technica
Myles McCormick and Claire Jones, Financial Times @ Ars Technica 1 place · today 10:40 EDT

US consumers, business pay 90% of tariff costs, says Federal Reserve

The Fed's research contradict Trump's claim foreign companies would bear the burden. Read more

316 fresh

Business Insider
Jake Epstein @ Business Insider 3 place · today 07:43 EDT

A US Navy destroyer and a support ship collided in the Caribbean during an at-sea resupply gone wrong

The incident, which occurred during a replenishment-at-sea, left two personnel injured. They're in stable condition. Read more

262 fresh

Business Insider
Chris Panella @ Business Insider · today 05:46 EDT

The US Air Force needs to buy hundreds of sixth-gen fighters and bombers to be ready for a China fight, airpower experts say

Without more aircraft to hit Chinese defenses, the Air Force might have to hold back, risking the military getting drawn into a slog, experts said. Read more

245 fresh

Engadget
Georgie Peru @ Engadget 2 place · today 11:15 EDT

Get two years of Proton VPN for 70 percent off right now

Proton VPN is offering a steep discount on its Proton VPN Plus subscription, with the two-year plan currently priced at $2.99 per month. You’ll pay $72 upfront for 24 months of service, which amounts to 70 percent off the usual monthly rate and brings the long-term cost well below what many premium VPNs typically charge.We’ve consistently been impressed by Proton VPN’s focus on privacy, its nonprofit ownership structure and the... Read more

243 fresh

The most popular news from the same source for the last week
Slashdot Slashdot
Slashdot
msmash @ Slashdot · 02/06/2026 10:34 EDT

The Bizarre Enhancement Claims Rocking Ski Jumping

German newspaper Bild reported in January that some ski jumpers have been injecting their penises with hyaluronic acid ahead of the Milan Cortina Winter Olympics -- the theory being that temporarily enlarged genitalia would yield looser-fitting suits when measured by 3D scanners, and those looser suits could act like sails to produce longer jumps. A study published last October in the scientific journal Frontiers found that a 2cm suit change... Read more

125

Slashdot
EditorDavid @ Slashdot · 02/07/2026 15:34 EDT

Good News: We Saved the Bees. Bad News: We Saved the Wrong Ones.

Despite urgent pleas to Americans to save the honeybees, "it was all based on a fallacy," writes Washington Post columnist Dana Milbank. "Honeybees were never in existential trouble. And well-meaning efforts to boost their numbers have accelerated the decline of native bees that actually are." "Suppose I were to say to you, 'I'm really worried about bird decline, so I've decided to take up keeping chickens.' You'd think I was... Read more

96

Slashdot
msmash @ Slashdot · 02/11/2026 01:00 EDT

The First Signs of Burnout Are Coming From the People Who Embrace AI the Most

An anonymous reader shares a report: The most seductive narrative in American work culture right now isn't that AI will take your job. It's that AI will save you from it. That's the version the industry has spent the last three years selling to millions of nervous people who are eager to buy it. Yes, some white-collar jobs will disappear. But for most other roles, the argument goes, AI is... Read more

93

Slashdot
msmash @ Slashdot · 02/09/2026 10:14 EDT

AI Gold Rush is Resurrecting China's Infamous 72-hour Work Week - in US

The AI boom has revived a workplace philosophy that China's own regulators cracked down on years ago: the 72-hour work week, known as 996 for its 9am-to-9pm, six-days-a-week cadence. US startups flush with venture capital are now openly advertising it as a feature, not a bug. Rilla, a New York-based AI company that monitors sales reps in the field, warns applicants on its careers page to expect roughly 70-hour weeks.... Read more

88

Slashdot
msmash @ Slashdot · 02/10/2026 12:00 EDT

A Bitcoin Blunder for the Ages: $40 Billion Accidentally Given Away

An anonymous reader shares a report: The hundreds of prize payouts were mostly just a few bucks each, part of a promotional campaign by a South Korean cryptocurrency exchange. The total reward pot: 620,000 Korean won, or about $425. Then came a colossal mistake. A staffer for Bithumb, South Korea's No. 2 crypto exchange, didn't distribute 620,000 Korean won. Rather, the prizes, due to an input error, emerged in a... Read more

81

Slashdot
msmash @ Slashdot · 02/09/2026 22:45 EDT

Electric Cars Are Making It Easier To Breathe, Study Finds

An anonymous reader shares a report: It turns out that when fewer cars spew exhaust as they drive along, air quality improves. That's the conclusion of a new study published in The Lancet Planetary Health that looked at the effect of increased numbers of both EVs and plug-in hybrids on air pollution in California. The Golden State has by far the largest number of plug-in vehicles in the United States,... Read more

72

Slashdot
EditorDavid @ Slashdot · 02/08/2026 18:34 EDT

Amazon Delivery Drone Crashes into Texas Apartment Building

"You can hear the hum of the drone," says a local newscaster, "but then the propellors come into contact with the building, chunks of the drone later seen falling down. The next video shows the drone on the ground, surrounded by smoke... "Amazon tells us there was minimal damage to the apartment building, adding they are working with the appropriate people to handle any repairs." But there were people standing... Read more

71

Slashdot
msmash @ Slashdot · 02/05/2026 15:00 EDT

Musk Predicts SpaceX Will Launch More AI Compute Per Year Than the Cumulative Total on Earth

Elon Musk told podcast host Dwarkesh Patel and Stripe co-founder John Collison that space will become the most economically compelling location for AI data centers in less than 36 months, a prediction rooted not in some exotic technical breakthrough but in the basic math of electricity supply: chip output is growing exponentially, and electrical output outside China is essentially flat. Solar panels in orbit generate roughly five times the power... Read more

63

Slashdot
EditorDavid @ Slashdot · 02/08/2026 21:34 EDT

Carmakers Rush To Remove Chinese Code Under New US Rules

"How Chinese is your car?" asks the Wall Street Journal. "Automakers are racing to work it out." Modern cars are packed with internet-connected widgets, many of them containing Chinese technology. Now, the car industry is scrambling to root out that tech ahead of a looming deadline, a test case for America's ability to decouple from Chinese supply chains. New U.S. rules will soon ban Chinese software in vehicle systems that... Read more

59

Slashdot
msmash @ Slashdot · 02/10/2026 16:00 EDT

Lost Soviet Moon Lander May Have Been Found

An anonymous reader shares a report: In 1966, a beach-ball-size robot bounced across the moon. Once it rolled to a stop, its four petal-like covers opened, exposing a camera that sent back the first picture taken on the surface of another world. This was Luna 9, the Soviet lander that was the earliest spacecraft to safely touchdown on the moon. While it paved the way toward interplanetary exploration, Luna 9's... Read more

56

Most popular sources

  • You see 932 news out of 937.
  • Sources 61 out of 61.
Business Insider 21% 1
Ars Technica 13% 2
Wired 9% 0
Gizmodo 7% 3
Tom's Hardware 5% 1
View sources »

LIKE us on Facebook so you won't miss the most important news of the day!

12.02.2026 12:47
Last update: 12:41 EDT.
News rating updated: 19:42.

What is Times42?

Times42 brings you the most popular news from tech news portals in real-time chart.
Read about us in FAQ section.


Times42 © 2026