Alphabet’s AI Advantage: DeepMind’s Wins and How It Could Still Fumble

When disruption comes around, the incumbents are usually caught off guard. With AI, that’s not the case. The leaders of the last decade are well positioned to be the leaders of the next, too. And, while there has been lots of speculation that the death of Google is near, the company continues to make strides to put that in the rear view mirror. Today we’re talking about the latest from Google’s DeepMind, and how Alphabet could fumble the bag.

Google This Week

This week, Google DeepMind released Veo 2, the company’s video generation model. The release came just one week after OpenAI’s highly anticipated Sora video generation model and, by many accounts, is much better.

Video generation models are still in their infancy, but with time, they could revolutionize content creation and entertainment. There are a number of companies looking to lead in this space, but Google’s DeepMind sits atop that pile as of mid-December with Veo 2.

Video generation isn’t the only thing making waves within Alphabet’s AI ranks this week though. There’s also robotics with DeepMind and Apptronik announcing a partnership.

“By combining Apptronik’s cutting-edge robotics platform with the Google DeepMind robotics team’s unparalleled AI expertise, we’re creating intelligent, versatile and safe robots that will transform industries and improve lives. United by a shared commitment to excellence, our two companies are poised to redefine the future of humanoid robotics.”

It’s not Alphabet’s first robotics play (the company previously owned Boston Dynamics), but it is the most interesting given the current climate.

A Humanoid Future

For a quick brief on Apptronik, this company creates humanoid robots to solve real-world problems. Things like working in a warehouse, unloading trucks, or tending to machines.

The obvious question you might ask is “why humanoid?” Why don’t we work on a new form factor? Well, that might come in time, but in 2024 it’s best to have these robots work alongside humans, and the tasks these robots are working on were designed around a bipedal with two arms and two legs.

By building humanoid robots, companies like Apptronik are building in a margin of safety to their venture. A distinct robot for unloading trucks might be more efficient, but it would require a lot more work and determination to bring it to market. You’d have to convince warehouses it’s faster, convince them it’s more efficient, and you’d have to get them to pay for that promise.

You know what’s easier to convince people of? That a robot with two arms and two legs can do the job humans were already doing. They can do it without risk of calling in sick, they can do it without being injured, and they can even be scaled up much easier than scaling up by hiring additional humans.

An Apptronik and DeepMind partnership brings the best of artificial intelligence at this time to these humanoid robots.

Apptronik will use Gemini’s multimodal capabilities to allow their robots to “understand” the world around them. This will allow for faster training, and working on more complex tasks. Gemini 2.0 touts the ability to “think, plan, and remember,” which will be a game changer in the robotics space.

Google’s Big Advantage

This piece opened by asking how Alphabet could fumble the bag. It is a legitimate question. Rarely do you see disruption coming and look around to find the disrupted perfectly positioned to win.

In Alphabet’s case, the company has a distinct hardware advantage. One might argue it’s not ahead in models, but it’d be tough to say they’re far behind. The biggest advantage the company has though is distribution.

Billions of Android handsets. Billions of daily Google searches. Over 1.5 billion active users of GMail. More than 720,000 hours of video footage uploaded to YouTube every day. Alphabet is a company with incomprehensible reach, and it’d have to stumble incredibly hard to lose out at this point.

OpenAI is a formidable opponent. The company has phenomenal name recognition with ChatGPT, it also has one of the best consumer available models out there with o1. To make matters worse for Google, OpenAI just made its search product free.

OpenAI is still going to struggle though. Without a moat, users will jump ship to a better UX, or for a better price. Alphabet can afford to, and will subsidize its victory over OpenAI.

That’s not to say OpenAI doesn’t have its place, it just doesn’t have an obvious play to disrupt Alphabet meaningfully. Its only play to do that, via Microsoft, is a window that’s rapidly closing as Microsoft looks to build ties with Anthropoic.

It’s worth a call out that this is all being written and published before OpenAI goes live with its 12th day of Christmas, so opinions may change a couple of hours from now.

Microsoft could disrupt, as could Meta or dark horse Apple, but those are the same competitors of decades past, and they don’t have the advantages stacked on their side like Alphabet. So that leads us to “how?” How can Alphabet fumble this bag?

Poor management and poor decisions. Last year the company released an image generation model that was so neutered it refused to generate pictures of white men. That’s poor decision making. Too much of that becomes poor management, and poor management when everyone’s trying to capture a piece of the pie leads to a fumble.

While Alphabet’s advantages in AI, robotics, and distribution seem insurmountable today, success is never guaranteed. The company’s scale and reach position it to shape the future of AI and robotics, but maintaining this dominance will require more than technological prowess. Alphabet must avoid complacency, ensure strategic agility, and balance innovation with ethical and operational discipline.

The tech landscape is littered with once-dominant players who faltered due to poor decisions, cultural inertia, or overconfidence. Alphabet’s greatest risk isn’t its competitors—it’s the temptation to rest on its laurels. As new entrants and old rivals push boundaries, Alphabet’s ability to stay hungry and focused will determine whether it continues to lead or becomes an incumbent left behind. For investors and observers, the next decade will reveal whether Alphabet’s moves today are enough to secure its future as the world’s AI and robotics powerhouse.