Perplexity wants to be the Meta, Google and Microsoft of the AI era

If you thought Perplexity's ambitions ended at winning in search, you haven't been paying close enough attention

This past week, two isolated pieces of news about Perplexity came out – one that was very high-profile, riding high on the heightened drama surrounding the TikTok ban (and unban, just hours later). The other flew more under-the-radar, a passing report about the startup exploring product direction decidedly outside of its current wheelhouse. Combined, they paint a picture of a company looking to grab as much real estate as possible during a pivotal paradigm shift.

The news you likely saw was that Perplexity made a bid for TikTok in the waning hours leading up to the ban; the news you may have missed is that Perplexity made an acquisition, of a professional social networking tool called ‘Read.cv’ that resembled a LinkedIn re-conceptualized from the ground up for the TikTok age.

This follows recent product moves by Perplexity to extend their capabilities in shopping, enterprise data lakes and collaboration, web publishing, podcast creation and more.

Given their initial launch product and direction, the general characterization of Perplexity as a potential Google-killer makes a lot of sense. And if it weren’t for the incredible (unprecedented?) entrenched advantage they have in terms of usage and spread, I think Perplexity’s product chops would allow it to outcompete Google’s search for a variety of uses – including much of the problem space of what search is actually intended to solve in the first place.

The focus on whether or not Perplexity is able to replace a juggernaut like Google, a bastion of the early internet age and the bulwark of the modern tech monopolists, distracts from more interesting discussions about what else Perplexity could become.

Supplanting legacy leaders in tech generally does not take the shape of a 1:1 replacement arriving and becoming a direct swap-in replacement for the incumbents. Instead, the vectors that challengers exploit are slightly off-kilter to the giants they hope to dethrone – basically, it’s the Death Star trench run (the Rebels never built a mirror Death Star of their own to take on the Empire, they tried to exploit a minor design flaw instead).

In the early days, Perplexity looked a lot like an OpenAI or Anthropic feature that neither of those two foundation model companies had gotten around to building yet. They managed to differentiate themselves even in the fundamentals to the point where it’s clear they have a product advantage even when compared directly to the OpenAI web search features that eventually launched. But even focusing on that is probably missing the forest for the trees.

Perplexity’s moves this past week reveal an opportunistic and questing curiosity when it comes to product direction and problem space expansion – this isn’t a company looking to rival Google in its place of power, but one that wants to find its way into fissures and gaps in Mountain View’s armor and pry open its defenses that way instead.

It’s still unclear whether the rise of generative AI will be a platform shift at the level of transitions like the move to desktop computing, the rise of the web, and the dominance of mobile. But, if it does reach that scale, the industry’s topographical changes are going to be shaped by moves made now, including some that happen on the periphery or without much fanfare.