AI’s need for a human touch

Hi everyone. The new wave of stories of humans pretending to be AI has a long history. But first...Three things you need to know today:• Ama [View in browser]( [Bloomberg]( [by Ellen Huet]( Hi everyone. The new wave of stories of humans pretending to be AI has a long history. But first... Three things you need to know today: • Amazon’s CEO says the AI boom [will be powered by his cloud division]( • The US warned government agencies [of a potential email breach]( • Taylor Swift’s songs [are back on TikTok]( Humans imitating computers imitating humans Recent headlines have made clear: If AI is doing an impressively good job at a human task, there’s a good chance that the task is actually being done by a human. When George Carlin’s estate sued the creators of a podcast who said they used AI to create a standup routine in the late comedian’s style, the podcasters claimed that the script had actually been [generated by a human]( named Chad. (The two sides recently settled the suit.) A company making AI-powered voice interfaces for fast-food drive-thrus [can only complete 30% of jobs]( without the help of a human reviewing its work. Amazon is [dropping]( its automated “Just Walk Out” checkout systems from new stores – a system that relied on [far more human verification]( than it was hoping for. We’ve seen this before – though it may already be lost to Silicon Valley’s pathologically short memory. Back in 2015, AI chatbots were the hot thing. Tech giants and startups alike pitched them as always-available, always-chipper, always-reliable assistants. One startup, x.ai, advertised an AI assistant who could read your emails and schedule your meetings. Another, GoButler, offered to book your flights or order your fries through a delivery app. Facebook also tested a do-anything concierge service called M, which could answer seemingly any question, do almost any task, and [draw you pictures]( on demand. But for all of those services, the “AI assistant” was often just a person. Back in 2016, I [wrote]( a story about this and interviewed workers whose job it was to be the human hiding behind the bot, making sure the bot never made a mistake or spoke nonsense. To power Facebook’s M, every single message was reviewed by contractors who worked out of the company’s Menlo Park headquarters. The workers at GoButler and [x.ai]( told me they worked long hours or overnight shifts to mimic the bots' perma-online presence. Customers constantly asked them if they were bot or human. Some customers sent them crude sexual messages, thinking no human would read it. (In another indicator of Silicon Valley’s rapid evolutionary cycle, x.ai closed in 2021, and the name is now attached to xAI, Elon Musk’s new AI venture.) People build AI to mimic human intelligence and capabilities. But when the AI can’t quite deliver on the promise, we end up with humans pretending to be chatbots pretending to be humans. It’s the latest iteration of a trick that stretches back at least as far as 1770, when the original Mechanical Turk machine appeared to play chess automatically — but actually concealed a human chessmaster inside its apparatus. Of course, some things are different this time around. Chatbots are more fluent in their conversation and mimicry. But today’s bots, just like 2016’s, struggle with reliability. We still trust a human more when something has to be done right. Investor frenzy is often driving this hype. In 2016, “Human-assisted AI is ‘the hottest space to be in right now,’” one founder said. Another founder said that she was frustrated with the way a lot of investors would “congratulate us on having a fully automated bot” when the technology was quite reliant on humans. As long as there’s the incentive to overhype AI’s abilities, there will be gaps between what AI promises and what it can reliably do. To fill that gap, you can always hire a person. —[Ellen Huet](mailto:ehuet4@bloomberg.net) The big story Apple plans to overhaul its Mac computers with its new more powerful M4 processor. The company hopes the chips will help revive sales on a product line that saw a [revenue decline of 27% in the past fiscal year](. Get fully charged Saudi Arabia and the UAE are [racing to build datacenters for AI](. Google alums are going big in Japan with a [startup that received a government grant of supercomputer time to work on artificial intelligence](. Reddit will take questions from users as part of its first [earnings conference call next month](. More from Bloomberg Bloomberg Technology Summit: Led by Bloomberg Businessweek Editor Brad Stone and Bloomberg TV Host and Executive Producer Emily Chang, this full-day experience in downtown San Francisco on May 9 brings together leading CEOs, tech visionaries and industry icons to explore the opportunities and pitfalls at the intersection of business and tech, from AI to the chip wars and beyond. With Evan Spiegel co-founder and CEO of Snap; Steve Huffman, co-founder & CEO of Reddit; Sarah Bond, president of XBox; and many others. [Learn More](. Get Bloomberg Tech weeklies in your inbox: - [Cyber Bulletin]( for coverage of the shadow world of hackers and cyber-espionage - [Game On]( for reporting on the video game business - [Power On]( for Apple scoops, consumer tech news and more - [Screentime]( for a front-row seat to the collision of Hollywood and Silicon Valley - [Soundbite]( for reporting on podcasting, the music industry and audio trends - [Q&AI]( for answers to all your questions about AI Follow Us Like getting this newsletter? [Subscribe to Bloomberg.com]( for unlimited access to trusted, data-driven journalism and subscriber-only insights. Want to sponsor this newsletter? [Get in touch here](. You received this message because you are subscribed to Bloomberg's Tech Daily newsletter. If a friend forwarded you this message, [sign up here]( to get it in your inbox. [Unsubscribe]( [Bloomberg.com]( [Contact Us]( Bloomberg L.P. 731 Lexington Avenue, New York, NY 10022 [Ads Powered By Liveintent]( [Ad Choices](