Hello, and welcome to TechScape. I’m your host, Blake Montgomery, and this week in tech news: Trump’s tariffs hit tech companies that move physical goods more than their digital-only counterparts. Two stories about AI’s effect on the labor market paint a murky picture. Meta released a standalone AI app, a product it claims already has a billion users through enforced omnipresence. OpenAI dialed back an obsequious version of ChatGPT. And we look back at Elon Musk’s first term.
Tech earnings: bits rake it in and atoms face uncertainty
Four of the magnificent seven tech giants reported their quarterly earnings last week. Meta, Microsoft, Apple and Amazon all beat Wall Street expectations, but their guidance for the coming quarters highlighted a divide between companies that ship physical products as opposed to digital ones, or atoms versus bits.
Meta and Microsoft’s earnings went gangbusters, beating expectations and offering strong guidance for the next quarter.
By contrast, clouds hung over Apple and Amazon. Both beat Wall Street’s estimates, but news about both last week highlighted the uncertainty and grim outlook brought on by Trump’s tariffs. At the end of Apple’s earnings call, CEO Tim Cook announced that the import taxes would cost the iPhone maker $900m in the next quarter alone. Apple can take the hit, and it has flown some $2bn worth of iPhones into the US from India in advance of the tariffs taking effect, but it’s an enormous number nonetheless.
Amazon faced the Trump administration’s direct wrath last week after Punchbowl News reported a rumor that the company would begin itemizing the cost added to the prices of individual items by the tariffs in the same way that discount retailers Shein and Temu have. White House press secretary Karoline Leavitt lashed out with a statement that displaying tariff line items on Amazon.com would constitute a “hostile and political act”. Amazon was quick to say that it had only considered the idea, never implemented it, and only with respect to its Shein and Temu competitor, Amazon Haul. After the dustup, the e-commerce giant announced it would not move forward with the idea.
Is AI eating up jobs?
Artificial intelligence (AI) is forecast to disrupt the labor market significantly. News stories have detailed affecting individual cases of jobs vaporizing and leaving employees in the lurch.
Tech skeptic Brian Merchant writes in his Blood in the Machine newsletter of Duolingo’s recent announcement that the language learning platform will become “AI-first” and “gradually stop using contractors to do work that AI can handle”. Merchant headlined his piece “The AI job crisis is here, now” and spoke to a former Duolingo contractor, a writer who said: “I did not expect to be replaced so soon.” Artists and illustrators likewise told Merchant that they had lost work to clients who said they would rely on AI instead.
At a macroscopic scale, though, we have yet to see turbulence predicted shortly after ChatGPT debuted. AI has been slower to upend the broader market than promised by the companies making it, according to new research. In a working paper released in April, economics researchers for the University of Chicago and the University of Copenhagen found that, in Denmark, “AI chatbots have had no significant impact on earnings or recorded hours in any occupation.” Rather than obviating whole jobs or even entire workday tasks, as has been foretold, AI has fit more into workers’ lives as a productivity tool, shaving off time from this or that item on the to-do list but also creating some new work. The researchers analyzed two large-scale adoption surveys from late 2023 and 2024 of 25,000 workers and 7,000 workplaces that covered 11 occupations thought to be threatened by AI.
Hat tip to the Register for its story on the paper.
I have never intentionally tried to have a conversation with Meta’s AI chatbot. I’ve only ever accidentally tapped the inconspicuous blue circle that appeared sometime in the spring of 2024 in my Instagram search bar. Doing so would prompt a new chat to open with the company’s AI agent – an all-too-eager chatbot that suggested I ask it to “imagine paradise” as a first prompt – instead of the list of my recent searches that I am accustomed to seeing. The company has made Meta AI difficult to avoid by placing it in frequently used parts of the company’s existing apps.
The search bar placement of Meta AI and takeover of Meta’s existing apps has been effective. On WhatsApp, for instance, a Meta AI button has also replaced the “new message” button in the bottom right corner of the iPhone app, making it even easier to accidentally tap. Since Meta first tweaked the search bars across all its platforms – Instagram, Facebook, Messenger and WhatsApp – with the option to fat-thumb your way into a conversation with a robot, the company has touted its fast-growing AI user base. For months, the company said Meta AI was “on track to become the most-used AI assistant in the world”. Most recently, the company said nearly 1 billion users use Meta AI.
Last week, Meta launched a standalone AI app, posing the question of whether its users want to interact with its AI if they don’t stumble into it. For now, the company expects most users to continue to encounter its AI via the blue circle placed conspicuously within its popular social apps, an executive told the Verge.
Meta is certainly not alone in this. Google includes an AI overview in a huge number of Google searches and yet similarly touts its over 1 billion AI overviews users (1.5 billion most recently). It’s hard to know how many of those users are willing participants, but it almost doesn’t matter. Intentional or not, the companies derive value from every interaction with their respective AI tools and make it near impossible for you to force Google or Meta to stop using your data to train their AI. In the US, for instance, you can only request that Meta delete your data or stop using it to train its AI. In addition to your chats with Meta AI, that can include your posts and other profile information.
There’s something dark about an AI reality that already is designed to feel inescapable so early into its life. Using these platforms in places like the US where there are few privacy regulations to protect you feels as if, whether you want to or not, you’re always training someone’s AI.
Sam Altman makes a mistake and a debut
OpenAI announced last week that it would roll back ChatGPT’s most recent update to the personality of its AI chatbot. “We missed the mark with last week’s GPT-4o update,” Sam Altman wrote. A week earlier, he had written, “The last couple of GPT-4o updates have made the personality too sycophant-y and annoying.”
after newsletter promotion
The update was rare misstep for the ChatGPT maker, which has been barreling forward and accruing users hand over fist ever since the bot’s debut in 2022. It boasts 400m weekly users, according to the venture capital firm Andreessen Horowitz, which has invested in OpenAI.
One day after the rollback, Altman announced the rollout to the US of World, his startup that manufactures an orb that scans a user’s iris as a means of verification. He crowed on X: “We did it!” with a picture of himself standing before an American flag that replaced the stars with his other company’s logo on a white background.
The Doge days
The richest man in the world and one of the loudest voices in tech occupied the White House for approximately 100 days. What has he done?
My colleague Nick Robins-Early writes:
“Musk has left little of the federal government unscathed. Over the course of just a few months, he has gutted agencies and public services that took decades to build while accumulating immense political power.
Musk’s role in the Trump administration is without modern precedent. Never before has the world’s richest person been deputized by the US president to cull the very agencies that oversee his businesses. Musk’s attempts to radically dismantle government bureaus have won him sprawling influence. His team has embedded its members in key roles across federal agencies, gained access to personal data on millions of Americans and fired tens of thousands of workers. SpaceX, where he is CEO, is now poised to take over potential government contracts worth billions. He has left a trail of chaos while seeding the government with his allies, who will probably help him profit and preserve his newfound power.”
Read more about Doge’s first 100 days.
