Keep It Simple, Genius (Premium)

OpenAI and Google faced off this week with dueling AI announcements, and the contrast couldn’t be more stark. I hope Microsoft and Apple are paying attention: Only one of these companies got it right. And it wasn’t Google.

I think we all understand that Google is in a tough spot. It gets almost zero credit for kicking off the AI era, and it has spent much of the past year in a defensive posture, trying to overcome perceptions and newly aggressive rivals that never seriously threatened it in the past. Complicating matters, the threat to Google is real: This company’s dominance is solely reliant on a single business—search and the resulting ad-based revenues it generates—that can and will be supplanted by AI. The only question now is where that AI will come from.

My in-the-moment hot take is that AI is like death by a thousand cuts, that there’s no single killer app but rather hundreds or thousands of individual advances across the entire spectrum of personal technology. And that each of us, each person, each business, will cherry-pick from the resulting matrix of advances and benefit from AI wherever we get it, wherever we need it most.

For this reason, Google is in a better position than many believe: It has over 15 businesses that each have over half a billion users—500 million users, each—giving it a scale that even Microsoft and Apple can’t match. And therein lies the problem: Google did mention this salient fact yesterday during its opening I/O 2024 keynote. But it came one hour and 51 minutes into a presentation that last one hour and 52 minutes. By that point, most of the audience had either stopped paying attention or had literally nodded off.

What preceded that stunning bit of clarity was the presentation version of a pummeling, a maddening list of advances that spans everything Google does, from low-level technologies like datacenter chipsets, LLMs, and developer APIs that span multiple platforms, often in contradictory ways, to end user apps like Gmail, Photos, and Maps … and everything in between. And that’s the problem: No matter how engaged and excited you are by these announcements, you’re only interested in some of it. And the interminable wait between items of interest was off-putting. Hell, Google saved its best Android news for the next day because there was too much to talk about.

The day before Google I/O kicked off, relative newcomer OpenAI gave free a masterclass in clarity and human awkwardness that Google—and now Microsoft and Apple—would do well to study and then emulate. Yes, I know. OpenAI is a newish company with a singular focus, and it has a lot less to discuss than does Google. But its Spring Update event was a crisp, non-boring 20 minutes or so, and any interested in diving deeper can head over to its YouTube channel and select from dozens of demo-based videos of GPT-4o, the most interesting announcement at that event.

As important, the OpenAI event was live, with all the glitches, mistakes, and awkwardness that that entails. And as strange as it is to write this, given the robot-like clinical nature of this company, that provided the much-needed humanity that’s often lacking at this type of event, and was specifically lacking in OpenAI’s public persona to date. This humanness was, I think, as important to making that event a success as was the technology it showed off.

Contrast this to the Google event, where cameras swept over the audience dramatically, but for no good reason because each presenter marched directly to an on-point circle on the stage and then stood there rigid and unmoving, as they robotically recited their scripts. (With some exceptions, of course. Dave Burke is always delightful, among others.) When you combine that with the sheer volume of information that Google unloaded, it was too much. Far too much.

So here’s my advice. Not as a presenter—I was never very good at that, frankly—but rather as someone who has spent the past 25 years or more sitting in the audience at industry events, much of it with Microsoft, suffering as people who should know better droned on and on about things that didn’t matter 10 seconds later let alone 10 years later. Say less. A lot less.

More specifically, what Google should have pummeled us with was 20 to 30 minutes of how AI is going to transform those 15 products with 500 million or more users, right up front, back-to-back. Followed by 10 to 15 minutes of how AI will transform Google’s platforms, Android and ChromeOS. Followed by 10 to 15 minutes of how AI will transform how developers create apps and services.

And then just end it. And follow that up with deeper-dive keynotes for each topic in turn so that interested parties can dive in and others can just get on with their lives. Don’t make everyone sit through a mind-numbing two hours of mostly superfluous (to each person) information, much of which went over their heads anyway. People tune out. You’re losing them.

I know this because Microsoft does this at its developer show every year. And it will almost certainly do it again next week, once again confusing volume with quality and computer science expertise with clarity. We’re at the point now where we need to understand what these companies are doing, but the key takeaway is only that they’re doing far too much, and we don’t understand almost any of it.

Google, Microsoft, Apple, and Amazon (let’s not forget about Amazon) are all in a position to lead, to continue their dominance in this new era. But they’re also on the precipice of being made irrelevant by upstarts like OpenAI and, more to the point, by this notion that AI will simply be everywhere, which means paying Big Tech to provide it may no longer be necessary. This is a moment in which those companies need to clearer, in their messaging and in the products they decide to release. But to date, they have mostly been obtuse, confusing us all with the sheer volume of what they’re doing. (Especially Google and Microsoft, two companies that are more alike than either is willing to admit.)

Guys, you’re losing us, and you’re doing so at a time when there are others waiting in the wings, eager to displace you. This is a chance to show leadership, but you’ve chosen an alternative path that I think is led by fear. Fear of the unknown. Fear of losing what you have.

Many years ago, I observed that liars tended to over-explain themselves and that the sheer volume of their words was an indication that the underlying message was bogus. Today, I’m applying this thinking to Big Tech and AI. Stop doing so much. Instead, act decisively, release the right products and updates, and then explain what you’re doing clearly.

I’m not optimistic that Microsoft will do much better next week, given history. And like most, I’m curious where Apple—which has its own AI perception problems—will land at WWDC in June, how it might leverage its own strengths, real and perceived, in this important year. We’re at a crossroads, and we’ll look back at 2023 and 2024 as when everything changed. Today, it’s still up in the air.

Good luck, Microsoft. You need it.