Tag: Microsoft

Will AI Ever Become a True Platform? Not Without Us….

Yesterday Microsoft CEO Satya Nadella published a fascinating essay that reads as both a rebuke to frontier AI model companies like Anthropic and OpenAI, as well as a call to arms to the entire business community.

Nadella is worried that OpenAI and Anthropic are about to eat everyone’s lunch – Microsoft included. “I’ve been thinking a lot about the future of the firm in an AI-driven economy,” he opines. “The last thing any of us want is a world where every company across every sector is ceding value to a few models that eat everything they see.”

Read More
Leave a comment on Will AI Ever Become a True Platform? Not Without Us….

SIGN UP FOR THE NEWSLETTER

Stay up to date on the latest from BattelleMedia.com

Should AI Be Addictive?

The most interesting piece of news this morning comes from Microsoft’s Satya Nadella, who made a very public point of chastising his own team for saying the quiet part out loud.

That quiet part? In an internal memo leaked to 404 Media, a Microsoft VP said his team’s goal was to “make people addicted” to Microsoft’s new Scout tool, which is fashioned on Open Claw, the AI agent project that went viral early this year. Nadella quickly quashed such sentiments, releasing a memo stating “this is absolutely a non goal! If anything we are doing the exact opposite. We want to make sure AI empowers and adds real value to human endeavor and broad economic growth! We should make our teams clear about this.”

Microsoft then went into damage control mode, with corporate comms chief Frank Shaw piling on that Scout is for “helping people accomplish tasks more effectively—not encouraging dependency. Our goal isn’t more screen time. It’s more time back.”

Read More
Leave a comment on Should AI Be Addictive?

Where’s All the AI Magic?

“Hey Google, how are the Giants doing this year?” 

I was standing at my bathroom sink, finishing up my daily ablutions, when a random thought popped into my head. It’s been a minute since I checked in on my favorite baseball team – ever since I moved East, they’ve been in a slump. Maybe they’re pulling back into contention this year? I still have a Google Home plugged in nearby. I’m intimately familiar with the Home’s limitations, but I asked anyway. Perhaps I was hoping one of my annual predictions (voice interfaces for the home) would magically come true.

Read More
1 Comment on Where’s All the AI Magic?

The Year Tech Gets Even Bigger (Predictions 2026, #7)

If only Nano Banana could spell, AI would be a thing.

For the past many years one prediction has proven reliably accurate: There will be no significant Federal regulation of the technology industry. At times this stalwart prognostication has been tested by major anti-trust actions – but each has proven ultimately toothless. This year, for example, we’ll learn what the DOJ managed to accomplish in its second case against Google – and it’s still possible a judge will rule that the search and AI giant must divest itself of its adtech infrastructure. But I don’t think so. And even if that ruling does come to pass, Google knows it can simply appeal, dragging out any eventual impact until it wins a war of attrition with an increasingly feckless and uninterested DOJ.

Besides, arguing about the past is playing yesterday’s game, and in 2026, the game has reverted to an even older playbook. For the past five or so years, tech giants have had to play defense when it comes to M&A and sweetheart partnerships – Meta was being sued over its acquisitions of Instagram and WhatsApp, Google over its consolidation of adtech and its domination of search distribution through deals with Apple and Samsung, among others. But in 2026, the governors are coming off.

Here’s what the co-head of software banking at Goldman has to say about the year ahead, according to The Information: “[Goldman] is gearing up for what it anticipates to be an “extremely active” period that could top 2021, a year where tech M&A and IPO volume both hit record-highs.” Put another way, Big Tech is gearing up to once again go big. Expect some eye-watering acquisitions, as well as novel partnerships around AI products and distribution. While the DOJ fights yesterday’s war, Google in particular will be building a Death Star of AI distribution featuring all the same players from its days of search monopoly: Apple, Samsung, and other Android partners. And the government won’t do a thing about it.

I opened last year’s predictions by noting that “Big Tech has leap-frogged finance in the pantheon of political influence,” and 2026 is the year that influence will start to pay its expected dividends. Yes, there will be individual actions driven by a mercurial and vengeful autocrat. But if tech companies continue to pay tribute, and I certain they will, that Goldman banker nailed it: 2026 is going to be a very, very big year for tech deals.

—

This is the fourth in a series of post I’ll be doing on predictions for 2026. The first three are here, here and here. When I get to #1, I’ll post a roundup like I usually do. 

You can follow whatever I’m doing next by signing up for my site newsletter here. Thanks for reading.

3 Comments on The Year Tech Gets Even Bigger (Predictions 2026, #7)

Is OpenAI Today’s Netscape? Or Is It AOL?

As is his want, last week Fred Wilson wrote a provocative post I’ve been thinking about for the past few days. Titled “Netscape and Microsoft Redux?“, Fred notes the parallels between the browser wars of the late 1990s and the present-day battle for dominance in the consumer AI market. And he asks a prescient question: What new, world-defining product might we be missing by focusing on AI chatbots?

In the early days of the Web, everyone thought the most important new product to emerge from the Internet was the browser. Netscape, a startup with just a few months of operating history, defined the market for those browsers in 1994, then dominated it for several years thereafter. But by the late 1990s, the lumbering incumbent Microsoft had stolen Netscape’s lead by leveraging distribution and pricing advantages inherent to its massive Windows monopoly.

But here’s the rub, as Fred points out: “Ironically, that battle for Internet dominance missed that the most important piece of software was the search engine, not the browser. And so the winner ended up being an entirely different company – Google.”

Fred notes that today’s version of the browser wars is playing out in chatbots, with OpenAI playing the role of the upstart (Netscape), and Google the incumbent (Microsoft). Sure, OpenAI has the lead today, but Google has woken up, and is using its dominance in search, infrastructure, and browsing to take share from its upstart competitor.

But if that metaphor holds, Fred wonders, are we once again missing “the most important new piece of software,” just as we did around search in the late ’90s? And if so, what is it?

I keep turning these questions over in my head, and it feels like an answer is tantalizingly close, but still out of reach. So I’m doing what I always do when faced with these kinds of puzzles: I’m thinking out loud through writing.

Let’s start by identifying what made search the breakout winner of the early Internet era. By the late 1990s, the dominant tool for accessing  the Internet was the browser. Those browsers let you surf an endless wave of sites, but they didn’t help you find the sites mattered to you.

The early Web had a navigation problem, and its first solution was the directory – Yahoo!, the first directory built for every day consumers, became the most visited site on the Web. But directories were soon swamped by waves of new sites coming online each month.

Google solved the Web’s navigation problem by continually crawling and indexing every site on the Internet, then delivering just the right answer based on a novel signal that had been overlooked by everyone: The link.

Google PageRank algorithm refined links – human-built connections between pages on the web –  into the most successful business in the history of the Internet, changing consumer behavior in the process. By the early 2000s, consumers began using Google as their first stop on the Web. Google became a verb, search became a habit, and portals like Yahoo languished.

Which leads me to wonder: If consumer AI really is in its “early 1990s” phase, what’s the equivalent of the navigation problem we all encountered prior to the emergence of Google search? And is there a novel signal – a new class of data emerging from our use of AI that corresponds to the link?

Exploration of these questions is complicated by how the Web works today. Twenty five years ago the Web was essentially a massive public commons. The vast majority of sites were open and free, and very few sites opted out of Google’s approach to crawling the Web. When people made new pages and links – essentially new data for Google to ingest – Google simply indexed that data, then used PageRank to figure out a way to make sense of it all. It didn’t have to ask permission – people gave it freely as a matter of course.*

Today’s Internet is markedly different. Most sites – in particular large platforms like Amazon, OpenAI, TikTok or Meta’s Instagram – operate as walled gardens that refuse to share data. And while people are still building websites, the majority of valuable data are locked behind walls of personalization and corporate terms of service. That means there’s no equivalent to the link in today’s AI-driven Internet – no novel public resource waiting to be exploited for a breakthrough service like search back in the day.

While we’ve not found an obvious analog to the link, perhaps I’ve been putting the horse before the cart. Let’s think for a spell about the problem. By the late 1990s, the problem of navigation on the Internet was widely understood. Is there a similar problem now for users of AI chat services?

The first thing that comes to mind is this: You can’t effectuate anything with services like ChatGPT or Gemini. You can’t ask them to take action on your behalf. Sure, you can ask them to write you an essay, act as a friend or therapist (or more), summarize a white paper, etc., etc., but as soon as you want to do something, chatbots grind to a halt.  Consumer AI has a “getting shit done” problem – they exist in rarified silos, incapable of anything that requires engagement beyond the confines of their chatbot interfaces.

Certainly the tech industry knows about this problem – and it has devised a solution: Agents. The next wave of AI innovation centers on “the agentic web,” with personalized agents that will do our bidding in every imaginable way. Every major AI company has announced agentic products, but unfortunately, they don’t work, because the ecosystem in which they operate is hostile to their success. Want an agent that compares prices across commerce sites like Amazon or Walmart, then makes a purchase? Sorry, ‘user agents’ are blocked by Amazon’s terms of service. In fact, nearly all commercial Internet services block ‘non human’ user agents from engaging with their sites. It’s not hard to understand why: Non human agents are, well, not human, and most sites depend on advertising revenue that’s targeted to humans, after all. Plus, user agents threaten to undermine the information asymmetry that underpins most of capitalism these days – once you’ve tasted the profits driven by dynamic pricing, it’s hard to go back.

What we have here is an architectural problem: The Internet as currently built simply cannot support an agentic future, no matter how many well-manicured hands might wave at it. Realizing that future will require a fundamental redesign of the Internet – a process as radical as the invention of the Web itself.

This leads me to an unexpected conclusion when it comes to pondering Fred’s timeline comparing the late 1990s to now. Perhaps, in fact, he’s off by a decade. Could it be that consumer AI is comparable not to the early Web, but rather to the era of nearly forgotten online services that preceded the early Web in the late 1980s? When the history of consumer AI is written, might it treat OpenAI, Claude and Gemini as the equivalent of pre-Web services like CompuServe, AOL, and MSN? These fascinating but frustrating services attempted to build online worlds, but they were built on brittle architectures that couldn’t connect easily and reliably to the “rest of the world.” The original Web offered a better model. Maybe it’s time to abandon the Web as we now know it, and dream up something entirely new once again.

—

*Plus, there was an explicit deal that quickly developed: Letting Google crawl your site meant inclusion in its index, resulting in visitors and potential business opportunities via advertising and/or commerce. 

—

You can follow whatever I’m doing next by signing up for my site newsletter here. Thanks for reading.

3 Comments on Is OpenAI Today’s Netscape? Or Is It AOL?

How Google Can Win the Future

[Second in a series, first post here]

This past week, Wall Street caught up with the rest of us and realized that Google has lost its monopoly grip on search. The trigger wasn’t Google losing an anti-trust case – that happened last summer. Nor was it the first ten days of Google’s ongoing search remedies trial. Instead, it was a statement just two days ago by an Apple executive, Eddie Cue, which led to an almost instantaneous panic amongst investors.

Cue told the court that consumers’ preference for using AI agents had led to a decline in search traffic inside Apple’s Safari browser (Google pays Apple more than $20 billion a year to secure that traffic – a major focal point of the government’s case).

Cue then twisted the knife: A decline in Safari’s Google traffic, he said, “has never happened in 22 years.”

A near perfect storm of “never happened” has settled over Google this past year: Twin rulings labeling the company a monopolist, a 10 percent annual decline in its stock compared to a 10 percent increase in the NASDAQ index, and the rise of generative AI as a replacement for search, the first significant competitive threat to the company’s 25-year run as the dominant force in how humanity asked questions of the world.

Poor Google. The company finds itself in the unenviable position of IBM and AT&T in the ’70s and ’80s, or Microsoft in the early 2000s. It’s become an industry titan beset by government antitrust actions that hobble its ability to respond to fundamental and permanent market changes.

What on earth can Google do? Well, I’ve got good news for the company: There’s a way out, but it’s going to hurt a little.

***

To understand the actions Google might take to ensure its continued membership in the Magnificent Seven, it’s worth taking a step back and studying Microsoft. Throughout the ’90s, Microsoft led the “Four Horsemen” – blue-chip tech stocks which drove unprecedented market returns during the first Internet boom (the four were Microsoft, Intel, Cisco, and Dell, if you can believe it). But by 1999, Microsoft was ruled a monopoly just as the center of the technology industry shifted from desktop PCs to a mobile-centric world built on top of the Internet.

For nearly two decades, Microsoft had dominated tech. Its Windows operating system captured north of 90 percent market share. Windows was the world’s #1 consumer-facing technology brand – the face of an industry. But once Microsoft settled with the US government and ended its five year anti-trust fight, the company’s brand, stock, and fortunes began a decade of decline. Founding CEO Bill Gates stepped down, and in a disastrous strategic misstep, newly installed CEO Steve Ballmer doubled down on the company’s past by putting Windows at the center of everything the company did. Here’s the Wall Street Journal’s summary of  the Ballmer era, a piece written as incoming CEO Satya Nadella was preparing to take the reins:

“In many cases, Microsoft latched onto technologies like smartphones, touchscreens, ‘smart’ cars and wristwatches … long before Apple or Google did. But it repeatedly killed promising projects if they threatened its cash cows [Windows and Office].”

Let’s update that paragraph, written in 2013, for present-day Google:

“In many cases, Google latched onto technologies like artificial intelligence long before Apple or OpenAI did. But it repeatedly killed promising projects if they threatened its cash cows [Search and Advertising].”

The lessons of Microsoft don’t end in 2013 – in fact, the company subsequently managed one of the most impressive turnarounds in modern business history. Acknowledging Ballmer’s mistakes, Nadella  deprioritized Windows and focused Microsoft on a decidedly less sexy but far more rewarding market: Enterprise infrastructure. By 2020, Microsoft was once again one of the most valuable tech companies in the world. Windows still existed, but as tech analyst Ben Thompson declared, it was no longer a driver of Microsoft’s business. Microsoft is now a leader in cloud-based infrastructure – an arms merchant selling highly profitable services to any and all comers.

***

Might Google learn from Microsoft’s stumbles and subsequent pivot? Indeed. As I hinted in my last post, it’s time for Google to release its death grip on consumer-facing search, and embrace a future where its cash cow becomes a commodity driving a new ecosystem of innovation. In that post, I wrote:

Google should make the database of intentions freely available to anyone who wishes to license it. 

In the remedies phase of Google’s search trial, which wraps today, the government has already demanded that Google share its core search data with any and all comers. Google has responded by taking to the fainting couch, declaring that such a remedy would “allow anyone to completely reverse engineer, end to end, every aspect of our technology stack.”

But would that really be so bad? At the end of my last post, I promised to lay out why Google sharing its search data could be a net win for not only Google, but the entire tech industry, and by extension, the world. Here’s my step by step guide:

Step 1. Negotiate a settlement with the government that embraces Google’s sharing of search data. If it acts quickly and keeps a long view, Google could negotiate the right to both charge market rates for its “search as a service” – which should bring in handsome profits – while continuing to offer its current search services (including Gemini AI). This would allow Google to continue to milk its consumer facing search application while also profiting from new applications built on top of its core search index and data.

Step 2. Design and launch a technology platform that delivers search as a service to qualified developers around the world. This platform would give partners access to Google’s search index in real time, and let them build whatever they chose to build on top of it. Like Microsoft, this would position Google as an arms merchant, selling high quality commodity search to all comers. Platform policy should include bi-directional data sharing, such that new services that leverage Google’s index would share engagement data back to the Google service, federating incremental improvements across all partners in the ecosystem.

Step 3. Boldly, loudly, and confidently begin to tell a new narrative about Google’s future.  Just as Microsoft was once a “Windows” company but moved past its history to become an enterprise winner, Google is evolving beyond search to become a neutral platform company that is accessible to all. It already has a robust asset to drive that journey in Google Cloud Platform, and this move will supercharge that business. Now the company will be single handedly responsible for unleashing a data asset – literally the database of intentions – that will create thousands of new companies and services, and with that tens of thousands of new jobs across the technology ecosystem.

Step 4. Continue to run and improve Google’s current search and AI services on top of this new platform. Wall Street hates abrupt change, and it’ll take years for AI and chatbots to eclipse traditional search in terms of revenue and profit. Don’t worry, Wall Street! Just as Microsoft kept selling Windows and Office, Google will keep selling search too.

Step 4. Imagine and build new services to compete on top of its own services. With consumer search – owning the customer at the interface layer – no longer the future of Google, Inc., the company will be free to compete with OpenAI, Perplexity, Kagi, and the thousands of new startups that will undoubtedly spring up once Google’s core search service becomes available as a platform. Instead of exhausting itself attempting to build a walled garden around search (and awkwardly shoehorning Gemini into its increasingly unwieldy core service), Google will be once again be a place where big ideas and crazy moonshots can be taken at scale.

Will this work? Of course there are obstacles, but I think it could. And it certainly sounds like a more appealing future than one of constantly defending yesterday’s business model of walled gardens and fealty to advertising economics. In one bold move, Google could unshackle itself from government oversight as well as a mirthless, spiraling death loop of enshittification.

In the end, Google turning search into a service would benefit the world in a way that would once again position the company as an innovator driving value across all of society – kind of like it was when I wrote The Search 20 years ago. Let’s face it: Google search has gotten pretty shitty over the past 20 years – and everyone’s noticed. If we’re headed into a world where AI chatbots become our interface to knowledge, the next generation of Claudes, Bings, and ChatGPTs will desperately require a high-quality search index that they, and we, can actually trust. That index could and should be Google’s. But only if the company has the courage to let go of its past.

—

Special thanks to the folks at Kagi, and in particular this blog post, which lays out the case for Google to share its database of intentions. 

You can follow whatever I’m doing next by signing up for my site newsletter here. Thanks for reading.

2 Comments on How Google Can Win the Future

The Question Google Won’t Answer

Reading Ben Thompson’s coverage of Google’s earnings call this week,  one thing jumps out, and simply can’t be ignored: Google CEO Sundar Pichai was asked a simple question, and, as Thompson points out, Pichai dodged it completely. A Merril analyst asked this question:

“Just wondering if you see any changes in query volumes, positive or negative, since you’ve seen the year evolve and more Search innovative experiences.”

Here’s Pichai’s answer:

“First of all, look, we think about effects on Search, obviously, more broadly. People have a lot of information choices. So — and user expectations are constantly evolving.

And so we’ve been doing this for a long time. And I think what ends up mattering is a strong continuous track record of innovation. Obviously, generative AI is a new tool in the arsenal. But there’s a lot more that goes into Search: the breadth, the depth the diversity across verticals, stability to follow through getting actually access to rich, diverse sources of content on the web, and putting it all together in a compelling way.

And I think, through the year two, when we test, we test Search Generative Experience, particularly against everything that’s out there. And we can see the progress we are making and how much users are liking the experience better. And so I think I feel very good about the progress.”

Not a yes, and not a no. Not a much of anything. More like an Olympic effort at contortion – Pichai was doing everything he could to avoid touching Google’s third rail: AI’s impact on its core cash cow of search. This tendency to avoid uncomfortable truths is understandable, but I think it’s going to come back to haunt the company. What I take away from it is that Google is not sure it will successfully manage the transition from traditional revenue models (AdWords mixed into old-school SERPs) to what its calling “Search Generative Experience” – where the company uses AI to deliver answers that mix various ingredients from the web to deliver a combination of useful information and, it hopes, commercial links that perform as well or better than traditional search monetization approaches.

Google is no longer “just a search company” – it has robust revenue lines with Workspace, Cloud, and YouTube, among others. But search still drives its core profitability, and search remains the product that will determine the company’s future path. At some point soon, Pichai will have to respond to the uncomfortable elephant in the room: Is search usage falling? And when he does, my guess is Wall Street won’t like his answer.

3 Comments on The Question Google Won’t Answer

Google Will Become the World’s Largest Subscription Service. Discuss.

A Google subscription box via Dall-E

Those of you who’ve been reading for a while may have noticed a break in my regular posts – it’s August, and that means vacation. I’ll be back at it after Labor Day, but an interesting story from The Information today is worth a brief note.

Titled How Google is Planning to Beat OpenAI, the piece details the progress of Google’s Gemini project, formed four months ago when the company merged its UK-based DeepMind unit with its Google Brain research group. Both groups were working on sophisticated AI projects, including LLMs, but with unique cultures, leadership, and code bases, they had little else in common. Alphabet CEO Sundar Pichai combined their efforts in an effort to speed his company’s time to market in the face of stiff competition from OpenAI and Microsoft.

Much is at stake. Google can’t afford to fall behind as its closest competitors throw massive resources at AI-driven products and services. But beyond keeping up, Google finds itself in an even higher-stakes transition: Its core business, search, may be shifting into an entirely new consumer model that threatens the very foundation of the company’s cash flow spigot: Advertising.

I’ve written about this topic quite a bit (here, here, and here), ultimately coming to the conclusion that over time, Google’s core business will shift from advertising to a direct-to-consumer subscription model. Generative AI chatbots are in the process of commoditizing Google’s core monopoly in search, and the race is on to create a new class of products that offer more value than the once-magical list of blue links. And because the chatbot interface doesn’t lend itself to the advertising model currently pouring hundreds of billions into Google’s coffers, the company will be forced to build something that regular folks like you and I will actually pay for.

This is why the stakes are so high for Google’s nascent Gemini project, which The Information reports will ship initial product (mainly for developers) this Fall. “Google is betting on Gemini to power services ranging from its Bard chatbot,” The Information writes,”which competes with OpenAI’s ChatGPT, to enterprise apps like Google Docs and Slides. Google also wants to charge app developers for access to Gemini through its Google Cloud server-rental unit.”

While the article doesn’t mention it, its reporting deepens my belief that in the long term, Google will adopt a business model that looks a lot like an ISP or mobile phone carrier: Charging one relatively high monthly subscription fee for access to a suite of services that nearly all consumers feel is mandatory to modern life. It may sound crazy to imagine hundreds of millions of us paying upwards of $100 a month for access to generative AI tools, but that’s because we’re using ChatGPT as a comp. Sure, ChatGPT is amazing, but it’s a point solution, not an integrated suite of services that feel essential to participation in modern society.

Once voice-driven chatbots are integrated throughout the Google universe, I can certainly imagine a universal Google subscription model coming into focus. Sure, you can use stripped down or ad-infested versions Google Docs, Mail, Search or YouTube for free, but a taste of those same applications when supercharged with voice-driven chatbot interfaces will lead most of us to conclude that we simply can’t live without a upgrade, and that will require a subscription. It may take a few years, but I’m more convinced than ever it’s coming. I’ll write more about why I feel this way in future posts – but for now, it’s back to the beach.

—

You can follow whatever I’m doing next by signing up for my site newsletter here. Thanks for reading.

3 Comments on Google Will Become the World’s Largest Subscription Service. Discuss.

Microsoft Ups the Ante on Both Google and Its Partner OpenAI

Microsoft today announced a cluster of upgrades to its Bing-ChatGPT product, including:

  • Eliminating the Bing chat waitlist, which effectively throttled the product’s growth by adding steps to a consumer’s journey.
  • Integrating more visual search results, which will enliven the consumer experience and potentially engage visitors for longer.
  • Adding chat history and persistence, a major differentiation between Bing chat and OpenAI’s ChatGPT, and for me anyway, the main reason I didn’t use Bing.
  • Adding more long document summarization, which is another feature that ChatGPT excels at.
  • Adding a platform layer to Bing, so third party developers can integrate in much the same manner as they can with ChatGPT’s plugins, which I’ve both praised and trashed in past posts (praised because of their potential, trashed because the model reminds me of the app store, which is a walled garden nightmare).

Overall, this news strikes me as Microsoft upping the ante not only on Google, which now has even more catching up to do, but also on Microsoft’s own partner OpenAI, which until now had a superior product. I’m on the road and not able to write as much as I’d like on this, but it’s worth noting. I’m sure the product managers in Mountain View aren’t getting much sleep these days – the pressure is mounting for Google to respond. And in OpenAI headquarters, the frustration has to be building as well – they cut that deal with Microsoft, and now have to live with its terms.

1 Comment on Microsoft Ups the Ante on Both Google and Its Partner OpenAI

AI Sausage-Making and Unconsidered Consequences

Is that an AI in your sausage?

Once again, Google and Microsoft are battling for the AI spotlight – this time with news around their offerings for developers and the enterprise*. These are decidedly less sexy markets – you won’t find breathless reports about the death of Google search this time around –  but they’re far more consequential, given their potential reach across the entire technology ecosystem.

Highlighting that consequence is Casey Newton’s recent scoop detailing layoffs impacting Microsoft’s “entire ethics and society team within the artificial intelligence organization.” This team was responsible for thinking independently about how Microsoft’s use of AI might create unintended negative consequences in the world. While the company continues to tout its investment in responsible AI** (as does every firm looking to make a profit in the field), Casey’s reporting raises serious questions, particularly given the Valley’s history of ignoring inconvenient truths.

In leaked audio that Casey reviewed, John Montgomery, Microsoft corporate vice president of AI, explains why the team was being disbanded: “The pressure from [CTO] Kevin [Scott] and [CEO] Satya [Nadella] is very very high to take these most recent openAI models and the ones that come after them and move them into customers hands at a very high speed.” Pressed by a staffer as to whether he’d reconsider, Montgomery responded: “I don’t think I will…Cause unfortunately the pressures remain the same. You don’t have the view that I have, and probably you can be thankful for that. There’s a lot of stuff being ground up into the sausage.”

A lot of stuff in the sausage, indeed. Montgomery was no doubt alluding to the idiom of seeing “how the sausage gets made” – a bloody mess involving parts of the animal most of us prefer to ignore (much less eat).

It may not be pretty, but it’s essential that society understand what’s going into the AI sausage – as well as the decision-making process behind how it gets made. And it’s also essential that companies making that sausage have internal controls independent of the processes (and payrolls) that favor profit and corporate advantage. From what I can tell from Casey’s reporting, it looks like that is no longer the case at Microsoft. The same seems to be true at Google, which famously mishandled the resignation/firing of Timnit Gebru, then the leader of its independent AI oversight team.

Losing independent oversight of corporate actors is scary, because we’ve been here before – over and over again. Remember the Cambridge Analytica scandal? For a brief moment after that mess, the Valley seemed united in realizing that rushing powerful, addictive, and at-scale technologies into the maws of powerful market forces might have some…unintended consequences. Directly after Cambridge, vocal critics Tristan Harris and Aza Raskin set up the well-funded Center for Humane Tech, and Facebook made a slew of promises, including “researching the impact of role of social media in elections, as well as democracy more generally.” It then set up its Oversight Board, the independence of which remains … questionable.

One of the most valuable lessons of Cambridge was a more general reassessment of risk as it relates to tech’s unintended consequences. “Unintended consequences can’t be eliminated, but we can get better at considering and mitigating them,” writes Rachel Botsman in a 2021 Wired piece Tech Leaders Can Do More to Avoid Unintended Consequences. Botsman was particularly concerned with the tech industry’s obsession with speed. “Speed is the enemy of trust,” she writes. “To make informed decisions about which products, services, people, and information deserve our trust, we need a bit of friction to slow us down.” She then quotes CHT co-founder Raskin: “If you can’t determine the impacts of the technology you’re about to unleash, it’s a sign you shouldn’t do it.”

If the companies most capable of unleashing AI aren’t identifying and fully exploring the unconsidered consequences of putting AI into everything, we’re speeding headlong into the same trap we did with Cambridge.

Which brings me to the statement at the top of this post – that Microsoft and Google’s enterprise and developer offerings are far more consequential than whether Bing steals a point or two of share from Google search. By offering ever-more powerful large-language models to developers at scale, both companies are unleashing a poorly considered and mostly unregulated technology to hundreds of thousands of developers and entrepreneurs, and by extension, to nearly every business on earth. It’s a gold rush mentality, the very same approach that gave us surveillance capitalism, Cambridge Analytica, and by extension, a widespread erosion of trust in democratic institutions.

Do we really want to run that play again?

—

*And yes, OpenAI announced ChatGPT4… 

**Microsoft objected to the possible conclusions that might be drawn from Casey’s story, and he appended their clarifications late yesterday. I’ll also add them here:

The story as written could leave the impression that Microsoft has de-invested broadly in responsible AI, which is not the case. The Ethics and Society team played a key role at the beginning of our responsible AI journey, incubating the culture of responsible innovation that Microsoft’s leadership is committed to. That initial work helped to spur the interdisciplinary way in which we work across research, policy, and engineering across Microsoft. Since 2017, we have worked hard to institutionalize this work and adopt organizational structures and governance processes that we know to be effective in integrating responsible AI considerations into our engineering systems and processes.

We have hundreds of people working on these issues across the company, including net new, dedicated responsible AI teams that have since been established and grown significantly during this time, including the Office of Responsible AI, and a responsible AI team known as RAIL that is embedded in the engineering team responsible for our Azure OpenAI Service. Less than ten team members were impacted on the Ethics and Society team and some of the former members now hold key positions within our Office of Responsible AI and the RAIL team.

 

Leave a comment on AI Sausage-Making and Unconsidered Consequences