The Breakout Tech Company Of 2023

I know readers are not going to love this choice. I apologize in advance. Is the company I’m about to pick below even a startup? What about highlighting the “little guy,” a new startup that’s quietly about to take off from the runway, a young company maybe everyone doesn’t know about yet? Or, “Hey, this isn’t the Heisman Trophy — you can’t win it twice, can you?!” And, who is this author, and how can he sit here for the past decade and just declare such things?

These are all fair questions. Historically, the litmus test for this annual choice has been a journey for me to identify one company of consequence which “broke out” from the insider chatter to more mainstream awareness. This year, as I struggled with how to convey this thought, I went back to a famous saying attributed to a baseball star from a bygone era. As part of the late 1970’s New York Yankees’ Bronx Bombers, Reggie Jackson never shied away from the ruthless NYC sports media. Instead, he welcomed the spotlight. “I’m the straw that stirs the drink,” Jackson famously said. The implication was that Reggie was the star, but more importantly, he set the agenda.

This brings me to this moment, where I select (again) OpenAI for the “Breakout Tech Company of 2023.”

Even as I write this as a startup junkie, a sucker for underdogs, a long-time early-stage investor, this 2023 journey is one that I have struggled with. There are countless startups that have indeed grown quickly over the past few years and are on the verge of scaling. As I debated these choices with friends and colleagues, however, there was never any clear consensus of who fit the bill for 2023. (A silver lining here is that the pump is primed for so many potential breakouts over the next few years.)

As I look back on my notes for 2023, Artificial Intelligence as a theme dominated technology (and policy) headlines, and it was again OpenAI who stirred the drink. In 2022, OpenAI broke out with a product that took off, ChatGPT; in 2023, OpenAI transformed into a  multi-use consumer-enterprise-etc platform. OpenAI’s “Developer Day” was heralded as, depending on your point of view, either a huge enterprise opportunity for developers (APIs, app stores, agents, and more!) or, perhaps, the death knell for scores of once-hopeful venture-fueled early-stage pre-product AI startups whose reasons for being were just demo-dayed away. OpenAI dominated headlines, sucking the oxygen away from many other deserving companies; this enabled the company to further position itself as a legitimate “big tech platform,” where they began to throw some weight around. OpenAI “gets the people going.”

But… Dangers lurk beneath.

While OpenAI has been winning a LOT, other players aren’t sitting around. Scores of technologists from the Google and OpenAI diasporas are hard at work building new and powerful platforms, such as Bard, Anthropic, or Adept, etc. Facebook can’t be forgotten here, with their formidable “Meta Research” organization’s LMaMA open source contribution, which sort of reminds me of Google’s (defensive) strategy with Android and the world of mobile operating systems in the mid-2000’s. It’s not just Facebook though, the world of open source believers (such as Stability, Mistral, and others) believe OpenAI is anything but “open,” betting their keystrokes that developers and enterprise customers will want to have more control of how AI is tuned and deployed within their own ecosystems and/or in front of their clients or consumers. Beyond the closed vs open debate, there’s active work on pushing models and compute further to the edge and on localizing models to specific verticals and/or geographies (for language reasons). There are also a few groups now working on what comes next after transformers, and perhaps we will learn more about them over the next few years.

Beyond different technological, architectural, and geographical considerations, exogenous issues related to AI remain unsettled — will the government over-regulate AI? Can incumbents layer in AI faster than upstarts can displace legacy companies? Could the “round-trip” revenues (and corresponding valuations) from underlying investors who are also customers present accounting challenges? And, with all the money being funneled into AI-related technologies, is the technology itself ultimately deflationary in nature, where the high-value variable costs today (think chips, servers, workloads, storage, compute, etc.) asymptote toward zero over time?

Those are all good questions, and I surely don’t have the answers. Neither does anyone, I’m sure, which is what makes OpenAI the pick for me for 2023. In a world awash in venture capital and interest in the next big thing in technology, there’s a lot of belief, but OpenAI is the player that alters the terrain when they announce a new release or spread their tentacles further into Microsoft, which owns nearly half of OpenAI (the corporate entity). There are a few big tech companies measured by trillion dollar market caps, and OpenAI is already swimming in these open waters, and if things continue for them as they have this year, they could be one of the biggest creatures in the sea. The straw that stirs the drink.

[I would like to thank my colleague Divya Dhulipala and old friend Josh Elman, both of whom provided lots of great fodder for debate on this topic.]