CEO of Forward Future, AI education platform and content creator
Matthew Berman
Profile
If you’ve spent any time in the AI corner of YouTube, you’ve almost certainly landed on a Matthew Berman video within hours of a new model dropping. That speed is his whole thesis. Berman is the founder and CEO of Forward Future, the education brand he launched in 2023 that now spans a YouTube channel of roughly 600K+ subscribers, a fast-growing daily newsletter, and a community aimed at helping developers and businesses actually use AI rather than just read headlines about it. He is not a researcher and doesn’t pretend to be — he’s an engineer-turned-communicator, San Francisco based, who before all this founded and ran Sonar Technologies for about nine years until its 2019 acquisition.
What Berman does well, and why he matters to someone learning AI, is triage. When Meta ships a new Llama, when DeepSeek surprises everyone, or when a fresh reasoning model appears, he publishes a hands-on walkthrough — usually within a day — running the model through a repeatable battery of tasks: write the game Snake in Python, solve a “killers in a room” logic puzzle, count the words in its own response, tackle a few reasoning and coding traps. It is not a rigorous benchmark, and he’d probably admit as much, but it’s a fast, consistent, apples-to-apples smell test that tells a busy developer whether a release is worth an afternoon of their attention. In a field where the signal-to-noise ratio is brutal, that consistency is the product.
The second thing Berman has quietly built is access. Forward Future’s interview slate has grown from testing local models on consumer GPUs into sit-downs with some of the most powerful people in tech — including Sundar Pichai of Google (twice, at Google I/O) and Satya Nadella of Microsoft. That’s an unusual trajectory for an independent creator, and it reflects how much the industry now values reaching the practitioner audience he’s assembled.
The honest caveat: Berman is an enthusiast, and his framing skews optimistic and fast-moving rather than skeptical or deep. His reviews are demonstrations, not evaluations, and his rubric can be gamed or simply not represent the workloads you care about. Treat his channel as a real-time radar for what’s new and roughly how capable it is — an excellent first filter — and pair it with hands-on testing and more technical sources when you’re actually choosing what to build on. For staying current without drowning, though, few do it more reliably.
Key Articles & Papers
Forward Future — The Future Today (Daily AI Newsletter) Forward Future on SubstackVideos
Controversies
No major scandals, but Berman has been candid about a real tension in the AI-content economy: he has publicly noted that negative, “AI-is-failing” content tends to get more views than measured coverage — an incentive he says he tries to resist rather than feed. Separately, in 2025 he banned generative AI from writing any of his newsletter’s prose, arguing that AI writing is “derivative by definition” and that originality is the one thing worth protecting — a notable stance from someone whose whole business is promoting the technology. Neither is a controversy so much as a window into the editorial pressures of being a full-time AI creator.
Spotify Podcasts
YouTube
