The AI Land Grab Begins

Instead of nightmare scenarios, let's start by scrutinizing how generative AI is going to change how we live today.

“Imagine you have a hammer. That's machine learning. It helped you climb a grueling mountain to reach the summit. That's machine learning's dominance of online data. On the mountaintop you find a vast pile of nails, cheaper than anything previously imaginable. That's the new smart sensor tech. An unbroken vista of virgin board stretches before you as far as you can see. That's the whole dumb world. Then you learn that any time you plant a nail in a board with your machine learning hammer, you can extract value from that formerly dumb plank. That's data monetization. What do you do? You start hammering like crazy and you never stop, unless somebody makes you stop. But there is nobody up here to make us stop. This is why the ‘internet of everything' is inevitable.”

--Shoshana Zuboff, The Age of Surveillance Capitalism, 2018

“The more of a focus we put on robots that are gonna come and kill us in our sleep, that's a nice distraction because what's actually happening in the meantime, there's a land grab that's happening right now.”

--Amy Webb, founder of The Future Today Institute and consultant to many Fortune 100 companies including Microsoft, speaking on This Week in Tech, May 15, 2023

“There’s no way a non-industry person can understand what’s possible. It's just too new, too hard, there's not the expertise. There's no one in the government who can get it right. But the industry can roughly get it right and then the government can put a regulatory structure around it."

--Former Google CEO and chairman Eric Schmidt, speaking on NBC’s Meet the Press, May 14, 2023

As we continue to muddle our way through all of the issues being raised by the explosive emergence of generative AI technologies, I think it’s worth holding three ideas in mind, which are nicely illuminated by the quotes above. First, as tech critic and philosopher Shoshana Zuboff explained so intensively in her book, left to its own logic, surveillance capitalism is going to keep eating the world. The second, from futurist and industry consultant, Amy Webb, reminds us that, for all the alarmist talk about an AI apocalypse, in the near-term we should pay attention to how companies like OpenAI (now a “long-term partner” of Microsoft) and Google are selling their AI services and disrupting existing economic models across society. And the third, from one of America’s more seemingly* public-spirited oligarchs, billionaire philanthropist Eric Schmidt, is a warning: Big Tech isn’t going to let government get in the way.

[*I say seemingly because Chairman Schmidt has a long history of obfuscation, going back to the early days at Google, when he develops its “hiding strategy” of not letting the public or regulators in on how it was mining user behavioral data to build its business. You may also recall how he tried to claim that retaining user data was inherent to the workings of search engines, in order to fend off early demands that old data be deleted (“the reality is that search engines including Google do retain this information for some time,” he claimed). Politico reporter Alex Johnson has done a bang-up job of unearthing the many ways Schmidt, who was very close to both Barack Obama and Hillary Clinton, has embedded his influence all over the Biden administration, and especially in areas touching on AI.]

What’s missing from this picture is how the rest of society responds to the rise of generative AI. That’s the wild card of the moment.

A lot of us are wondering right now how AI will change politics. That was the title of my friend Katie Harbath’s Substack post from a week ago. In it, she offered a set of potential tech innovations and pondered their potential good and bad outcomes. The innovations she explored included text generation (using tools like ChatGPT to create campaign content), image/video/audio generation, data analysis (to better find voters and persuade them), issue analysis (to help voters understand where candidates stand), and voter planning (to help people figure out what’s on their ballot and where to vote). It’s a good list for starters, but immediately also shows the problem with trying to get your head around the AI challenge.

Politics, after all, is about much more than what campaigns do to gain votes or how voters figure out who to vote for and where. Formally speaking, that’s just what elections are about. And though it’s still very early days in the deployment of AI at societal scale, it’s not hard to see some tectonic shifts coming.

The first, I suspect, is now at our doorstep, with Google’s announcements last week about how it is going to be “supercharging search with generative AI.” The examples the company showcased were all related to personal lifestyle decisions, like shopping choices, planning a family vacation, or setting up an exercise plan. But they all point toward the same result: giving a user a satisfactorily comprehensive answer to their question. As the company put it in a blog post:

“When searching for a product, you’ll get a snapshot of noteworthy factors to consider and products that fit the bill. You’ll also get product descriptions that include relevant, up-to-date reviews, ratings, prices and product images. That’s because this new generative AI shopping experience is built on Google’s Shopping Graph, which has more than 35 billion product listings — making it the world’s most comprehensive dataset of constantly-changing products, sellers, brands, reviews and inventory out there. In fact, every hour, more than 1.8 billion listings are refreshed in our Shopping Graph to give people fresh, reliable results.”

If you’re in the advocacy or publishing business, why should this concern you? Here’s why. “Supercharging” Google Search with AI is going to choke your web traffic. In 2020, two-thirds of searches on Google ended in search results without a user clicking forward onto another website, according to a SimilarWeb analysis of about 5.1 trillion Google searches on a panel of more than 100 million mobile and desktop devices that SimilarWeb tracks. That’s an increase in a similarly reported no-click-through rate in 2019. Over 87% of all internet searches in the US are done via Google, so this is no small matter. By adding more instant answers, knowledge boxes and other kinds of personalized explanatory information with generative AI, Google is going to retain even more user attention—without compensating publishers a penny for providing all their valuable content.

Because this is what we do now, I asked GTP4 about this and here's what it said:

“If users are indeed satisfied with the detailed and personalized answers provided directly by Google's AI-enhanced search results, it might lead to a decrease in organic traffic to other websites. This could have several implications for organizations, publishers, and campaigns that rely heavily on search engine traffic:

1. Reduced visibility: With fewer links appearing in search results, websites may experience a drop in their visibility, making it more challenging for them to reach their target audience.

2. Decreased ad revenue: Publishers that generate revenue from on-site advertisements might see a decline in earnings if there is a significant drop in incoming traffic from search engines.

3. Increased importance of alternative traffic sources: Organizations, publishers, and campaigns may need to diversify their traffic sources, placing more emphasis on social media marketing, email marketing, and other online advertising channels to compensate for the decrease in search traffic.

4. More focus on user experience: To retain visitors and encourage return visits, websites might need to focus more on providing an exceptional user experience, including high-quality content, helpful resources, and seamless navigation.

5. Greater dependence on paid advertising: Some organizations might turn to paid search engine advertising, such as Google Ads, to maintain their online presence and attract users to their websites.

6. Increased competition for ranking: The reduced number of links in search results could lead to fiercer competition among websites to be included in those results, making SEO strategies more crucial than ever.In summary, AI integration in Google search results could lead to a significant shift in the digital landscape, forcing organizations, publishers, and campaigns to adapt and find alternative ways to reach their target audience and maintain visibility online.”

Laura Hazard Owen has a good round-up on the Nieman Lab site digging into the ramifications of this change for publishers, and there’s plenty of speculation about whether this shift, which Google is just starting to roll out, will work to the benefit of large publishers over smaller ones. Arguably, if you have a big email or SMS list that you use for reaching your audience or supporters, this shift won’t affect you much. But should you make big investments in improving your website’s user experience or put more effort into cultivating influencers or social media marketing? Maybe? Your mileage may vary, as they say. But this is just one illustration of a larger point: even seemingly well-intentioned deployments of AI are going to alter settled patterns of behavior.

Back, briefly, to the wild card: how we respond. A decade ago, Google set its sights on building a next generation “smart city,” one of several ambitious efforts its leaders dreamt up to satisfy co-founder Larry Page’s desire for “moonshot” level projects. Back then the hot buzzword was “big data,” remember? Then-CEO/chairman Eric Schmidt recruited former New York City deputy mayor Dan Doctoroff to start a Google subsidiary called Sidewalk Labs and together they zeroed in on Toronto as their petri dish. As reporter Josh O’Kane details in his terrific 2022 book Sideways: The City That Google Couldn’t Buy, at first the project seemed completely wired for success. But civic activists and veteran city planners, plus a few independent Canadian tech moguls, started raising questions, especially as Doctoroff’s hubris and arrogance began rubbing people the wrong way. “What he wanted was to run his own government,” smart city expert Anthony Townsend, who worked for Sidewalk Labs in its early days, told O’Kane.

Indeed—in addition to seeking to levy and collect its own taxes, Sidewalk wanted to radically expand how it would collect and use public data for its own for-profit goals. An in-depth vision document called the “Yellow Book” laid out a horrifically revealing set of plans for how Google would build “a city from the internet up,” including such wild ideas as installing “smart mirrors” in people’s homes so the city could use hidden sensors to monitor their faces for signs of increased stress; charging people under a “pay-what-you-throw” system in order to incentivize them to produce less garbage; and tiered access for residents to their own neighborhoods based on how much data about themselves they would be willing to share. But here’s the point: Public engagement in public oversight processes slowed Sidewalk down and eventually stopped it in its tracks. Sidewalk Labs is dead. The future is not only what our technological overlords want it to be. So if you don’t want the future of AI regulation to be set by a handful of companies and their billionaire investors, start asking questions and in particular, demand that your own representatives—local, state and national—pay attention.

—Related: Say hello to Data & Society’s new Algorithmic Impact Methods Lab, which will develop public systems for evaluating how automated decision-making systems are impacting people’s lives and society at large.

—Also related: More than 150 content moderators for third-party outsourcing companies based in Africa that do work for Facebook, TikTok and ChatGPT have voted to form the African Content Moderators Union, Billy Perrigo reports for Time magazine.

Odds and Ends

—The Washington Post’s data visualization team has created “Tinder for budget nerds.”

—Code for America is facing complaints from its employee union that it is engaging in bad-faith bargaining.

—If you live in New York City and are at least 11 years old, you can participate in The People’s Money, the city’s first city-wide participatory budgeting process. Only $5 million is at stake, however, a far cry from the $500 million the Civic Engagement Commission originally called for. (Individual council members already have the option of setting aside $1 million from their own capital funding allotments for participatory budgeting by district.)

—Taking campaign finance grifting to a whole new level.

Deep Thoughts

Mark Engler and Paul Engler have written a comprehensive study for Waging Nonviolence of Barcelona en Comú, a popular movement that took over the city eight years ago and tried to create a model of a “fearless city” that sought to channel mass popular participation into a new way of doing politics centered less on party factions and more on direct listening and collaborating. “Perhaps most notable,” they write, “is the online Decidim platform, through which more than 100,000 registered users have voted on citizen-generated proposals for neighborhood improvements and engaged in participatory budgeting processes that, in 2022, distributed some 30 million euros in resources.” But the movement has faced tremendous opposition from entrenched power and has found many of its biggest initiative stymied. If you’re an alternative futures nerd, the Englers have done you a service.