OpenAI and Google kicked off the week by releasing competing multimodal artificial intelligence products. On Monday, OpenAI announced its flagship model ChatGPT-4o — a multimodal chatbot that can see, hear, and have real-time conversations. On its heels, Google unveiled a prototype of its AI assistant that it says…

Read more...

As usual, Google I/O 2024 is an absolute whirlwind of news and announcements. This year, rather than focusing on hardware, Android, or Chrome, Google spent most of this year’s developers’ conference convincing us that its AI features are worth prioritizing. One of those projects is Project Astra, a multimodal AI…

Read more...

Sony Music suspects Google, Microsoft and OpenAI have used music from the likes of Adele and Beyonce.

This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.

It’s common, in criminal court, for a defendant’s friends and family to join them in the courtroom as a show of love and support. That’s not exactly what’s happening in Manhattan this week. More, after these three stories from The Atlantic:

Michael Schuman: China has gotten the trade war it deserves. The art of survival The Baby Reindeer mess was inevitable.

Trump’s Courtroom Groupies

Donald Trump’s hush-money trial was already strange enough. A former president and a porn star walk into a Manhattan courtroom. But an additional cast of characters have recently inserted themselves into the drama. During this week’s testimony from Trump’s former attorney Michael Cohen, Republican politicians of many different ranks donned their courtroom best and headed downtown to put on a show for the boss. Although these particular charges could be the weakest of the many indictments Trump faces, one got the sense that none of his party allies was there to discuss the finer points of the law.

Speaker of the House Mike Johnson, who quite famously opposes pornography on religious grounds, nevertheless accompanied the man accused of cheating on his wife with a porn star to his trial yesterday. Outside the building, Johnson told reporters that the case is a “sham” and a “ridiculous prosecution.” At Monday’s session, Senators Tommy Tuberville of Alabama and J. D. Vance of Ohio apparently adopted the roles of Tom Servo and Crow T. Robot for the event, each offering running commentary on the “dingy” courtroom, with its “depressing” vibes; the disturbing number of mask-wearing attendees; and the “psychological torture” being inflicted on Trump (who, as a reminder, has been charged with 34 felonies in this case alone).

The former presidential candidates Vivek Ramaswamy and North Dakota Governor Doug Burgum took turns bemoaning the state of the justice system. Even a few bit players wanted in on the action, including pro-Trump Iowa Attorney General Brenna Bird, who flew all the way from Des Moines to stand before the news cameras and remind viewers that “politics has no place in a criminal prosecution.”

You might be wondering: Don’t these people have anything better to do? The answer is that, in today’s Republican Party, prostration before Trump is as much part of the job as anything else.

All of this showboating has been happening for several reasons. “At its most tangible level, what we’re seeing is a work-around to the gag order,” Sarah Isgur, a former senior spokesperson for the Trump-era Justice Department, told me. The ex-president was warned by the presiding judge, Juan Merchan, that if he talks or posts any more about the jury or the judge’s family, he could face jail time. So, just as in a political campaign, Trump’s surrogates are stepping up to stump for him.

Legally, Trump can’t ask these politicians to violate the order on his behalf. But why would he have to ask when they know exactly what he wants from them? “The playbook that Trump expects party members in good standing to follow is in public,” Amanda Carpenter, a former GOP staffer and now an editor at Protect Democracy, told me. Trump’s given these acolytes their cues with his posts on Truth Social, and they’re dutifully following them.

Lower Manhattan has, over the past week, become a pilgrimage site for those vying to be in Trump’s inner circle, with a court appearance carrying the promise of a holy anointing. You could also think about this courtside display as another audition in the early veepstakes. For Ramaswamy, Burgum, and Vance, in particular, this moment is a chance to demonstrate their abiding loyalty to Trump in the hopes of being selected as his running mate. Others may have reasoned that a little time in front of the cameras yelling “Sham trial!” will go a long way toward snagging a plum Cabinet position in a second Trump administration. (A virtual unknown such as Bird, of course, probably doesn’t expect to get either of these things. But, as every climber knows, one must never miss an opportunity to ingratiate oneself with the boss.)

Which brings us to one final observation. “There’s a larger, more philosophical reason they’re all there,” Isgur said. “That’s what the Republican Party stands for now.” There is no real platform, no consistent set of principles. There is only Trump, and degrees of loyalty to him. These courtroom groupies are simply responding to the obvious incentives—“If you didn’t know that the Republican Party is now focused on Trump,” Isgur said, “I’ve got an oceanfront property to sell you in Arizona.”

That’s their deal, but what about the boss’s? Trump no longer appears concerned only with shielding himself from political accountability. Now that he has almost clinched the nomination, he’s using the party to shield himself from criminal accountability, too. This has given the GOP a new rallying cry. “The Big Lie in 2020 was that the election was stolen,” Carpenter said. “The Big Lie 2.0 is that justice has been weaponized against him to deprive him of the presidency.”

Related:

Trump’s alternate-reality criminal trial What Donald Trump fears most

Today’s News

Slovakia’s prime minister, Robert Fico, was shot several times in an assassination attempt. He has been hospitalized and is undergoing surgery. President Joe Biden and Donald Trump agreed to two presidential debates. The first one will be hosted by CNN on June 27, and the second will take place on September 10, broadcast by ABC. A barge slammed into the Pelican Island Causeway in Galveston, Texas, causing a partial collapse and spilling oil in the bay below.

Dispatches

Atlantic Intelligence: This week, Google and OpenAI announced competing visions for the future of generative AI, Matteo Wong writes.

Explore all of our newsletters here.

Evening Read

Photo-illustration by The Atlantic. Source: Alamy.

The New Workplace Power Symbols

By Michael Waters

If you walked into an office building during the second half of the 20th century, you could probably figure out who had power with a single glance: Just look for the person in the corner office. The corner offices of yore were big, with large windows offering city views and constant streams of light, plus unbeatable levels of privacy. Everyone wanted them, but only those at the top got them. Land in one, and you’d know you’d made it.

Fast-forward to today, and that emblem of corporate success is dying off. The number of private offices along the side of a building, a category that includes those in the corner, has shrunk by about half since 2021, according to the real-estate company CBRE. But today’s workplace transformation goes beyond the corner.

Read the full article.

More From The Atlantic

What Alice Munro has left us The horseshoe theory of Google Search The eight dynamics that will shape the election

Culture Break

Timothy A. Clary / AFP / Getty

Look. The dogs at the 148th Westminster Kennel Club Dog Show made a big splash. Check out the images from this year’s competition.

Read. “Tapering,” a poem by Jane Huffman:

“I’m tapering / the doctor says–– // It might feel / like you can hear / your eyes moving––”

Play our daily crossword.

P.S.

For the past few months, it wasn’t exactly clear whether President Biden and Trump would meet on the debate stage before November. (Some, including The Atlantic’s David Frum, argued that they shouldn’t.) Well, folks, it looks like they’re doing it. The two presidential contenders agreed today to participate in two debates. The agreed-upon rules stipulate that neither event will have a studio audience—a welcome development for viewers who would rather watch a political debate than an episode of Jerry Springer.

— Elaine

Stephanie Bai contributed to this newsletter.

When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic.

OpenAI and Google showcased their latest and greatest AI technology this week. For the last two years, tech companies have raced to make AI models smarter, but now a new focus has emerged: make them multimodal. OpenAI and Google are zeroing in on AI that can seamlessly switch between its robotic mouth, eyes, and ears.

Read more...

This is Atlantic Intelligence, a limited-run series in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.

This week has felt like the early days of the generative-AI boom, filled with dazzling events concerning the future of the technology.

On Monday, OpenAI held a last-minute “Spring Update” event in which the company announced its newest AI model, GPT-4o, in an impressive live demo. Running on the iPhone’s ChatGPT app, the model appeared able to understand live camera footage, help solve a math problem, and translate a live conversation between English and Italian speakers. Every previous smartphone assistant, including Apple’s Siri, now appears obsolete—and the smartphone itself might be reimagined as the device most “perfectly positioned to run generative-AI programs,” as I wrote on Monday.

Not to be outdone, Google followed suit yesterday during its annual developer conference, which focused almost exclusively on generative AI. Alongside various technological advances, the company laid out its vision for search in the AI era: “Google will do the Googling for you,” Liz Reid, the company’s head of search, declared. In other words, chatbots will take the work out of finding information online. Lurking beneath the announcements was an acknowledgment that AI is better suited at synthesizing existing information and formatting it into an accessible format than providing clear, definitive answers. “​​That’s not omniscience; it’s the ability to tap into Google’s preexisting index of the web,” I noted yesterday.

OpenAI’s and Google’s announcements are a duel over not just which product is best, but which kind of generative AI will be most useful to people. The ChatGPT app promises to do everything, all in one place; AI-powered Google search promises to be more of an open-ended guide. Whether users will embrace either vision of a remade web remains to be seen.

— Matteo Wong, associate editor

What to Read Next

The live translation exhibited in OpenAI’s demo gestures toward a world in which people no longer feel compelled to learn foreign languages. Louise Matsakis explored this potential AI-induced death of bilingualism in March, writing, “We may find that we’ve allowed deep human connections to be replaced by communication that’s technically proficient but ultimately hollow.” Another, perhaps even more alarming future for the world’s languages: Generative AI, which is most proficient in the handful of languages with plentiful training data, may push thousands of other tongues into extinction. “If generative AI indeed becomes the portal through which the internet is accessed,” I wrote in March, “then billions of people may in fact be worse off than they are today.”

P.S.

The buzz over AI image generators and deepfakes was preceded, decades ago, by similar excitement and hand-wringing over Adobe’s Photoshop, then “the primary battlefield for debates around fake imagery,” my colleague Caroline Mimbs Nyce wrote last week. In the age of AI, Photoshop is struggling to adjust to being just one player in a crowded field.

— Matteo

The demonstration had no noticeable impact on the conference attendance or on the enthusiasm of the crowd inside the venue.

If you walked into an office building during the second half of the 20th century, you could probably figure out who had power with a single glance: Just look for the person in the corner office. The corner offices of yore were big, with large windows offering city views and constant streams of light, plus unbeatable levels of privacy. Everyone wanted them, but only those at the top got them. Land in one, and you’d know you’d made it.  

Fast-forward to today, and that emblem of corporate success is dying off. The number of private offices along the side of a building, a category that includes those in the corner, has shrunk by about half since 2021, according to the real-estate company CBRE. But today’s workplace transformation goes beyond the corner. All assigned desks and offices are on the decline, comprising only 45 percent of the average office, compared with 56 percent in 2021. Employers of many kinds—law firms, oil companies, biotech businesses, railway operators—are doing without them. They’re being replaced with coffee areas, conference rooms, and other collaborative spaces, Janet Pogue, the global director of workplace research at Gensler, a design firm, told me. In 2023, communal areas constituted 20 percent of the average office floor, up from 14 percent in 2021, CBRE found.  

[Read: Corporate buzzwords are how workers pretend to be adults]

Some take this shift as evidence of a revolution in egalitarianism. But tearing down the walls of the corner office hasn’t exactly made the workplace more democratic. Executives still claim many of the new shared spaces as their own, take the lead during team meetings, and enjoy other subtle privileges. They may no longer occupy a seat of such visible power, but they still exert sway over the rest of the office, using spaces as they please and influencing what others can do in them.

The relationship between office geography and power has always been in flux. The earliest workplaces displayed a “total inversion of our image of the traditional office-space hierarchy,” Dale Bradley, a professor at Brock University who has researched the history of the office, told me. Take the Larkin Building, which was designed in 1904 for a soap company. There, executives sat in the center of the first floor. “These guys are on display all the time,” Bradley said. Lower-level employees, meanwhile, were pushed to the edges of the building.

But as the years passed, bosses decided they’d rather not be so visible. Shortly before World War II, they started moving into the corners, where they could easily block themselves off, and the symbol of corporate power that we know today took root. By mid-century, developers were bragging about how many corner offices they could cram onto each floor. In 1966, The New York Times wrote about a Philadelphia office building that had eight corners—“twice the normal supply.” Toronto’s Scotia Plaza did even better, with as many as 22 on some floors. This influence seeped into literature too. Books with titles such as Nice Girls Don’t Get the Corner Office and Winners Dream: A Journey From Corner Store to Corner Office made clear that attaining the corner office was the apex of capitalist triumph.  

Then, about two decades ago, office designers began phasing them out. Google’s open-plan office redesign in 2005 was one of the earlier and most high-profile signs of changing attitudes: Modern workplaces were emphasizing collaboration, and corner offices were seen as too siloed. In 2020, the office came under attack again. Especially during the early pandemic, fewer executives were showing up to work in person, and their palatial corner offices were going to waste. “Nobody else really feels comfortable going into somebody else’s space and sitting there,” Kay Sargent, a director at the design firm HOK, told me. Now new office leases are about 20 percent smaller than they were before COVID, and companies are reconfiguring their floor plans to include more communal areas so that more employees fit into less space.

Without the corner office, status is conveyed in new ways. No matter the setup, “human beings will still find a way of creating hierarchy,” Lenny Beaudoin, CBRE’s global head of workplace design, explained. Bosses might have more computer monitors, bigger desks, or even just a permanent spot rather than a rotating one, Matthew Davis, a business professor at Leeds University Business School, told me. Power also manifests intangibly—for instance, only a select few might be able to not check Slack or come and go from the office without explanation. It’s the same benefit of having a far-flung corner office, re-created digitally: You know you’re important if you can escape surveillance.

[Read: ‘Resimercial:’ the terrible word for today’s trendy office aesthetic]

And even if they’re not in the corner, a lot of executives do still have offices. Those have largely slimmed down, but many are connected to conference rooms or other collaborative spaces, such as broadcast rooms in finance firms, recording studios at media companies, and labs in the life sciences. Many higher-ups essentially seize these for themselves whenever they come in, Pogue, at Gensler, told me. From there, they can shape any collaboration that takes place, ensuring it plays out in their space and under their supervision. Many modern companies “have as many conference rooms as there are executives,” Sargent said, and it’s become a “dirty little secret” that conference rooms are the new corner offices.

A collaborative space doesn’t need to be directly attached to a boss’s office for this dynamic to play out. When a high-ranking executive parks themselves in a big conference room or spreads their stuff across the long table in the office coffee shop, no one is going to tell them to leave. The communal spaces are free to use only if the boss isn’t there.   

Across all iterations of the workplace, higher-ups have always found ways to get what they want. We may one day return to an older layout of explicit hierarchy: Beaudoin, the CBRE designer, told me that he recently worked with a bank that decided to reinstall corner offices for a group of senior leaders, hoping it would bring them back to the office. But even if recent changes prove lasting, with space designed to be up for grabs, there won’t be any illusions about who has the power to grab it.

Earlier today, Google presented a new vision for its flagship search engine, one that is uniquely tailored to the generative-AI moment. With advanced technology at its disposal, “Google will do the Googling for you,” Liz Reid, the company’s head of search, declared onstage at the company’s annual software conference.

Googling something rarely yields an immediate, definitive answer. You enter a query, confront a wall of blue links, open a zillion tabs, and wade through them to find the most relevant information. If that doesn’t work, you refine the search and start again. Now Google is rolling out “AI overviews” that might compile a map of “anniversary worthy” restaurants in Dallas sorted by ambiance (live music, rooftop patios, and the like), comb recipe websites to create meal plans, structure an introduction to an unfamiliar topic, and so on.

The various other generative-AI features shown today—code-writing tools, a new image-generating model, assistants for Google Workspace and Android phones—were buoyed by the usual claims about how AI will be able to automate or assist you with any task. But laced throughout the announcements seemed to be a veiled admission of generative AI’s shortcomings: The technology is great at synthesizing and recontextualizing information. It’s not the best at giving definitive answers. Perhaps as a result, the company seems to be hoping that generative AI can turn its search bar into a sort of educational aid—a tool to guide your inquiry rather than fully resolving it on its own.

This mission was made explicit in the company’s introduction of LearnLM, a suite of AI models that will be integrated into Google Search, the stand-alone Gemini chatbot, and YouTube. You will soon be able to ask Gemini to make a “Simpler” search overview or “Break It Down” into digestible chunks, and to ask questions in the middle of academic YouTube videos such as recorded lectures. AI tools that can teach any subject, or explain any scientific paper, are also in the works. “Generative AI enables you to have an interactive experience with information that allows you to then imbibe it better,” Ben Gomes, the senior vice president of learning and the longtime head of search at Google, told me in an interview yesterday.

The obvious, immediate question that LearnLM, and Google’s entire suite of AI products, raises is: Why would anybody trust this technology to reliably plan their wedding anniversary, let alone teach their child? Generative AI is infamous for making things up and then authoritatively asserting them as truth. Google’s very first generative-AI demo involved such an error, sending the company’s stock cratering by 9 percent. Hallucinate, the term used when an AI model invents things, was Dictionary.com’s 2023 word of the year. Last month, the tech columnist Geoffrey Fowler pilloried Google’s AI-powered-search experiment as a product that “makes up facts, misinterprets questions, delivers out-of-date information and just generally blathers on.” Needless to say, an SAT tutor who occasionally hallucinates that the square root of 16 is five will not be an SAT tutor for long.

[Read: The tragedy of Google Search]

There are, in fairness, a plethora of techniques that Google and other companies use in an attempt to ground AI outputs in established facts. Google and Bing searches that use AI provide long lists of footnotes and links (although these host their own share of scams and unreliable sources). But the search giant’s announcements today, and my interview yesterday, suggest that the company is resolving these problems in part by reframing the role of AI altogether. As Gomes told me, generative AI can serve as a “learning companion,” a technology that can “stimulate curiosity” rather than deliver one final answer.

The LearnLM models, Gomes said, are being designed to point people to outside sources, so they can get “information from multiple perspectives” and “verify in multiple places that this is exactly what you want.” The LearnLM tools can simplify and help explain concepts in a dialogue, but they are not designed to be arbiters of truth. Rather, Gomes wants the AI to push people toward the educators and creators that already exist on the internet. “That’s the best way of building trust,” he said.

[Read: What to do about the junkification of the internet]

This strategy extends to the other AI features Google is bringing to search too. The AI overviews, Gomes told me, “rely heavily on pointing you back to web resources for you to be able to verify that the information is correct.” Google’s three unique advantages over competing products, Reid said at the conference, are its access to real-time information, advanced ranking algorithms, and Gemini. The majority of Google Search’s value, in other words, has nothing to do with generative AI; instead, it comes from the information online that Google can already pull up, and which a chatbot can simply translate into a digestible format. Again and again, the conference returned to Gemini’s access to the highest-quality real-time information. That’s not omniscience; it’s the ability to tap into Google’s preexisting index of the web.

That’s arguably what generative AI is best designed for. These algorithms are trained to find statistical patterns and predict words in a sentence, not discern fact from falsehood. That makes them potentially great at linking unrelated ideas, simplifying concepts, devising mnemonics, or pointing users to other content on the web. Every AI overview is “complete with a range of perspectives and links to dive deeper,” Reid said—that is, a better-formatted and more relevant version of the wall of blue links that Google has served for decades.

Generative AI, then, is in some ways providing a return to what Google Search was before the company infused it with product marketing and snippets and sidebars and Wikipedia extracts—all of which have arguably contributed to the degradation of the product. The AI-powered searches that Google executives described didn’t seem like going to an oracle so much as a more pleasant version of Google: pulling together the relevant tabs, pointing you to the most useful links, and perhaps even encouraging you to click on them.

Maybe Gemini can help sort through the keyword-stuffed junk that has afflicted the search engine. Certainly, that is the purpose of the educational AI that Gomes told me about. A chatbot, in this more humble form, will streamline but not upend the work of searching, and in turn learning.

Google is preemptively hedging against any risk of backlash from creative industries over its new AI media-generation tools — by bringing artists in as spokespeople.

Read more...