Remember when the height of presentation craft was a star wipe transition and a clip-art handshake? Somebody spent an afternoon picking a font, dropped in six bullet points, and called it a deck. The format hasn’t really changed since. The tooling got prettier, the templates got tasteful, the export options multiplied, and we all still sit in rooms watching someone read bullets off a wall.
That’s the backdrop for Gamma’s acquisition of Lica, an Accel-backed design startup, announced in 2026. Terms were undisclosed. Lica’s co-founders will run a new AI design research lab inside Gamma, aimed at new presentation methods. Lica’s product turned screenshots and recordings into presentations and videos.
That’s the whole factual footprint. Which, honestly, is most of what I want to talk about.
What this actually is
An acqui-hire with a mission statement. I don’t mean that as an insult. Gamma didn’t buy a customer base or a moat, it bought two founders and pointed them at a problem. The output of the deal isn’t a feature ship date, it’s a team charter. “Design research lab” is the tell. Labs are where you put people when you want them thinking about the shape of the thing rather than the roadmap for the thing.
For a company whose entire pitch is generating decks, that’s a defensible bet. Gamma’s core product does the same job as every other AI presentation tool on the market, and that job is essentially a formatting job. Type a prompt, receive slides. The differentiation window on that is narrow and closing, because the underlying models do the heavy lifting and everyone rents the same models. When your product category is one prompt deep, the only real edge left is deciding what the artifact should be in the first place.
Why Lica specifically makes sense
Lica wasn’t a slide company. It took screenshots and screen recordings and turned them into presentations and videos. That’s a different starting assumption. It treats the source material as messy artifacts of actual work rather than an empty canvas waiting for a title slide. Video output matters too, because it drops the pretense that a human will be standing there narrating.
If Gamma wants to move past slides, hiring people who already built something that wasn’t slides is more sensible than asking your existing team to unlearn the product they’ve been shipping. Internal teams optimize what they’re measured on. A lab exists specifically to not do that.
Where my skepticism kicks in
Research labs inside product companies have a rough survival rate. The pattern is familiar: acquisition gets announced, founders get titles, and eighteen months later the “lab” has been quietly folded into the core team and its people are fixing export bugs. The pressure is structural. A lab produces uncertain output on an uncertain timeline while the main product needs shipping every quarter.
A few things I’d watch for, in rough order of how much they’d tell you:
- Whether the lab ships anything that looks unlike a slide deck. If everything that comes out of it is a nicer deck, the charter didn’t survive contact with the roadmap.
- Whether Lica’s co-founders are still there in a year. Founder retention after an acqui-hire is the cheapest signal you get.
- Whether the video and recording-to-output pieces show up in Gamma’s main product, or vanish. Absorbed capability is a real outcome. Abandoned capability is a real outcome too.
- Whether “adapts to the audience” turns into something you can point at, or stays a phrase in a launch post.
The undisclosed terms don’t tell us much either way. Small acqui-hires usually go undisclosed because the number isn’t flattering to anyone. That’s not a knock on Lica. Most design startups in a category dominated by model providers end up choosing between a slow grind and a soft landing with people who’ll actually use what they built.
The part that’s genuinely interesting
Underneath the corporate-development noise, there’s a real question here. Presentation software has spent forty years assuming a human, a room, and a linear sequence of static frames. Almost none of that is true anymore. Decks get skimmed in a browser, forwarded without context, summarized by a model before anyone reads them, and consumed asynchronously by people who will never hear you talk.
If a lab actually attacks that assumption, the result should look strange. Documents that reorganize depending on who opened them. Output that’s a two-minute video for one reader and a dense table for another. Formats that don’t have a first slide because there’s no beginning to a thing nobody presents.
Gamma has bought itself permission to try. Whether it uses that permission or spends it on prettier templates is the only thing about this deal that will matter in a year.
🕒 Published: