Good morning. And hey, be honest: Do I look cute in this pic?
A new Washington Post report details how some people are asking AI chatbots whether they’re attractive or not. Then they’re asking the chatbots to—gulp—recommend remedies.
Some users engage in the practice with the perception that the AI is objective. (We know it’s not.) Others do it to get a second opinion…or perhaps a Simon Cowell-style roast.
But with chatbots reflecting the data they’re trained on—the flawed data of the flawed humans who generated it—I can’t help but wonder if there isn’t a better disclaimer than “it can make mistakes.” Because something like this isn’t a matter of accuracy.
By the way: In the interest of science, I uploaded a photo of myself to three different chatbots. All outwardly professed to keep things respectful and constructive; only two dared to assess me—and offered Style 101 generalisms as solutions. Your job’s safe for now, Cowell. —Andrew Nusca
Want to send thoughts or suggestions to Coins2Day Tech? Drop a line here.
Apple CEO criticized after skipping Trump's Middle East trip
President Trump had hoped to get some of the biggest U.S. Tech CEOs to join him on his recent jaunt to the Middle East, where he announced big plans for AI investments.
Tim Cook didn’t bite, according to a new New York Times report—and it didn’t take long before the Apple chief executive drew Trump’s public ire.
“I mean, Tim Cook isn’t here but you are,” Trump said to Nvidia CEO Jensen Huang at a Riyadh event attended by OpenAI’s Sam Altman and AMD’s Lisa Su.
Later, in Qatar, Trump mentioned that he “had a little problem with Tim Cook,” criticizing Apple’s investments in expanding its assembly operations in India.
By the end of the week, Trump published a social media post threatening tariffs of 25% on iPhones made anywhere except the United States.
Cook deftly managed his relationship with Trump through the president’s first term. Trump described the Apple CEO as a “great executive” even as Cook pushed back on some of Trump’s policies.
Things aren’t going so well during Trump’s second term.
Can Cook find middle ground when the President of the United States wants to immediately reshore the single biggest business Apple has? Stay tuned. —AN
EU regulators say Shein isn’t treating shoppers properly
The hits keep coming for Chinese e-commerce giant Shein.
Still licking its wounds from the outbreak of the China-U.S. Trade war, Shein is now in the crosshairs of European Union regulators who say the company’s sales tactics run afoul of the law.
EU watchdogs say Shein has violated rules against false discounts, deceptive product labels, misleading information, and website designs that pressure shoppers to complete purchases.
It’s also in hot water for sustainability claims that are allegedly misleading and a customer service department that’s allegedly difficult to reach. (Shein said it’s committed to compliance.)
Like the U.S., the EU has considered imposing fees on low-value packages imported into the bloc—a step that could hurt the low-cost business models of retailers Shein, Temu, and others.
EU regulators are also concerned about the safety of items from the companies as well as the lack of data about them.
Shein has about a month to respond to the regulators; if it fails to address their concerns, it could be fined. —AN
Meta’s AI challenge: Reverse the brain drain
Fun fact: Just three of the 14 authors credited on the pioneering 2023 research paper that introduced Meta’s “Llama” AI models remain at the company.
That revelation comes courtesy of a Business Insider report about the brain drain underway at the parent company of Facebook, Instagram, and WhatsApp.
According to BI, research scientist Hugo Touvron, research engineer Xavier Martinet, and technical program leader Faisal Azhar still retain Meta IDs.
But their colleagues have left for AI rivals like Mistral (the Paris startup partly founded by Meta vets Guillaume Lample and Timothée Lacroix); Anthropic; Cohere; Google DeepMind; Kyutai; Microsoft; and Mira Murati’s Thinking Machines Lab.
Meta has laid off thousands of employees in a bid to focus its efforts on the AI race, spending eye-watering amounts of capital—as much as $72 billion this year alone—on technical infrastructure and talent.
But there’s plenty of work still to be done.
The head of the Fundamental AI Research Group (FAIR) at Meta, Joelle Pineau, said last month that she would step down. Llama 4, the company’s latest AI release, has been poorly received. And Meta’s largest-ever AI model, dubbed “Behemoth,” is reportedly delayed.
Meta CEO Mark Zuckerberg disrupted his company’s operations to go all in on AI last year; now he just needs to create stability to allow people to get to work. —AN
More tech
—A U.S.-Japan sovereign fund? SoftBank’s Masayoshi Son made his pitch for U.S. Tech infrastructure.
—Voice AI attracts dollars. VC investment in voice AI startups serving call centers rose 7X, to $2.1 billion, in two years’ time.
—OpenAI makes moves in South Korea. Legal entity? Check. Seoul office? Double check.
—JioHotstar ascendant. The subscriber base for the India streamer, partly owned by Disney, now totals 280 million, driven by pro cricket.
—Meituan beats estimates. The Beijing food delivery giant reported Q1 revenue up 18.1%, to about $12.1 billion, as it looks beyond China.
—Apple’s new “Solarium” interface. The likely centerpiece of the company’s upcoming developer conference.
—Singapore “scamdemic.” Scam victims in the city-state lost $858 million last year, up 70% from the year prior.
—Xiaomi’s self-developed chip. It’s finally here and called the Xring O1.
Endstop triggered
