Exclusive: Gamma ravamps presentation platform to make output look more human
SiliconANGLE Paul Gillin
Gamma rebuilt its presentation app to make AI slides look less generic and plug into company tools. It’s a bet that “less AI-looking” output is now a product feature, not a bonus.
Based on reporting by SiliconANGLE, Paul Gillin — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Gamma has pushed out a rebuilt version of its presentation platform, Gamma 5, and the company’s target is pretty clear: make AI-made slides look less like AI made them. The new release adds a redesigned editor, a stronger AI agent and a much larger library of visual styles, all aimed at breaking the sameness that has started to make AI presentations feel instantly recognizable.
Co-founder and chief product officer Jon Noronha said users have been complaining more often this year about the “AI smell” in slides — the repetitive layouts, the generic imagery, the telltale design choices. That’s a blunt phrase, but a useful one. Gamma seems to have heard it as a product requirement, not just a branding problem.
The rebuilt editor gives users more freedom than earlier versions, which kept them inside a narrower set of design components. Now the agent can write custom code, draw shapes and arrange elements on a canvas. It runs in a sandbox where it can make images, work with files, do web research and check its own output, while automated checks catch problems like text running off a slide or awkwardly cropped pictures. Users can still step in at the element level with a freeform editor, and they can work with the agent on an outline before any slides are generated.
Gamma is also leaning hard into company context. It has around 20 connectors for pulling in outside information, including Slack, Granola, Fireflies and Fathom, and it can import PowerPoint files, PDFs and other existing material for editing. The company says those imports have gotten far better, with an internal visual-accuracy benchmark moving from roughly 30% early in development to over 97%. Brand preferences, custom instructions and workspace templates now carry across projects, so teams can keep reusing the same look and context.
Under the hood, Gamma says its platform uses more than 20 AI models, including models from Anthropic, OpenAI and Google, plus several image generators. The company picks models through internal tests and user experiments, watching whether people edit, share and pay. Founded in 2020, Gamma says it has had about 110 million sign-ups for its free product and has raised $90 million from investors including Andreessen Horowitz and Accel. Noronha also said some users cut presentation creation from eight to 10 hours to about 30 minutes, mainly by avoiding formatting work.
My take — AI-written commentary, not fact-checked reporting
Gamma is reading the room better than most AI product teams. People do not want “magic” if the magic leaves them with slides that look like they were assembled by a polite robot in a hurry. The real race now is against sameness, and that’s a far more honest market than hype about replacing designers ever was.
Read more about this at: SiliconANGLE