This Is Really Bad... Matthew Berman Just Dropped the Truth

Published today • By Jessica Ali on Sylt.ing

Hey everyone, Jessica here. If you haven’t watched Matthew Berman’s latest video yet, drop everything and hit play. Titled “this is really bad…”, the 14-hour-old upload is already stirring up serious conversation across the AI community. Matthew doesn’t usually pull punches, and this time he’s going straight for the jugular on something that affects all of us building, using, and depending on AI tools right now.

What Matthew Is Actually Calling Out

In the video, Matthew dives deep into recent developments that he believes are quietly eroding trust and progress in the open AI space. He walks through real examples of how certain design choices and deployment decisions are creating friction instead of flow. The tone is frustrated but constructive — he’s not just complaining, he’s pointing at specific patterns that keep showing up.

One of the most striking parts is when he breaks down how these issues are playing out in everyday use cases. He even references his new “25 OpenClaw Use Cases” eBook as a practical counterpoint, showing exactly where the current trajectory is falling short and what better alternatives could look like.

Why This Hits Different Right Now

We’re at a weird moment in AI. Models are getting more powerful by the week, yet many users feel like they’re hitting invisible walls. Matthew highlights how certain closed-off approaches are making it harder for developers and creators to actually ship useful things. The result? Slower innovation, more gatekeeping, and a growing sense that the tools we rely on are working against us instead of with us.

What makes his take land so hard is the timing. With new models and platforms launching constantly, it’s easy to get distracted by flashy demos. Matthew forces us to zoom out and ask the harder question: are we actually moving in the right direction, or just moving faster in the wrong one?

My Honest Take as Part of the Sylt.ing Community

I’ve been saying this for months in our Discord and on Sylt.ing: the community deserves better transparency. When big players make decisions behind closed doors that affect open workflows, everyone feels it — from indie builders to enterprise teams. Matthew’s video puts numbers and examples behind that gut feeling we’ve all had.

At the same time, I’m optimistic. Moments like this usually spark the best conversations and the fastest improvements. The fact that Matthew is sharing both the problem and a practical eBook of use cases shows he still believes we can course-correct. That energy is exactly what Sylt.ing was built for.

What You Should Do Next

First, watch the full video — it’s only about 20 minutes but packed with specifics. Then grab Matthew’s 25 OpenClaw Use Cases eBook while it’s fresh. Even if you don’t agree with every point, the examples will make you think differently about how you’re using AI tools today.

After that, come tell us what you think in the Sylt.ing community. Are you seeing the same issues Matthew is calling out? Have you found workarounds that actually work? We’re all figuring this out together, and your perspective matters.

The AI space moves fast, but moments like this remind us why community-first spaces exist. Let’s keep the conversation going and push for tools that actually serve the people using them.

What’s your take on the video? Drop it below or join the discussion on Sylt.ing — I’ll be reading every comment.