Stop Trying to Pick the Winner
Jack Dorsey's Buzz is a free, open-source Slack-and-GitHub killer with AI agents baked in. I'm running it, and here's why picking a winner is the wrong game, plus the test I use instead.
Yesterday Jack Dorsey launched a tool that lets AI agents join your team like actual coworkers. It's called Buzz, it comes from Block, and it points straight at Slack and GitHub at the same time. The announcement did 2.5 million views in a day.
It's free. It's open-source. Every agent you add gets its own cryptographic identity and an audit trail, so when a bot merges code or approves a workflow, there's a paper trail back to a human. That's a real architectural difference from bolting a chatbot into Slack, and it's a sharp idea.
I've got it running on my machine right now. And I'm still not migrating my company to it, at least not yet.
Here's why that restraint is the right call. And it has nothing to do with Buzz being early.
Everyone's asking the same question about every one of these launches: is this the one? Is this the Slack killer? Buzz, Viktor, Paperclip, Claude, Centaur. Pick your fighter. It's the wrong question.
Because in a market this young, the winner is temporary by design.
Two things guarantee it. First, half these tools are open-source now, which means there's no moat to defend. Second, and this is the part people miss, the big labs are watching. Anthropic and OpenAI are sitting there taking notes on what actually works, then folding it straight into ChatGPT, Claude, and Codex. The OpenClaws and Hermeses of the world speed up a desktop workflow, it catches on, and six months later the same feature ships inside the official app for free. Whatever "wins" this year gets swallowed next year.
So picking a winner is playing the wrong game. The right game is picking what pays you back this month, and staying disloyal on purpose.
I already run this playbook, because I have to. We're bootstrapped. Every dollar we don't spend is a lifesaver, and every automation that actually earns its keep is a little more superpower to go get the next thing done.
So I keep a running audit of what we overpay for. We dumped Intercom for Chatwoot and saved over $500 a month. We're into HubSpot for twenty grand a year. Slack is fifteen bucks a head, every month, forever. Jira. On and on. It adds up to a number that would make any bootstrapper sweat.
A free, open-source tool that does chat and code hosting and native AI agents in one place? That's not a curiosity. That's a threat to three line items at once. I don't need Buzz to win. I need it to make one of those bills negotiable.
That's the shift I want you to make. Stop reading these launches as "should I switch?" Start reading them as "what does this let me stop paying for?"
Which brings me to the actual system. When a tool like Buzz drops, I don't agonize. I run it through four gut-checks. I call it the Four-Shoe Test, because you've got to try every shoe on to find the one that fits your team.
1. Does it replace a bill? This is first for a reason. If a tool could kill or shrink something I already pay for, it gets a look immediately. Chatwoot passed this test against Intercom and saved us $6,000 a year. A tool that only adds a shiny new capability goes to the back of the line. A tool that removes cost jumps the queue.
2. One workflow, one week. I never migrate a team. I give the new tool exactly one real workflow (a sprint, a client activation, one channel) for one week. Real work is the only honest test. If it survives contact with an actual deadline, it earns more rope. If everyone drifts back to the old tool out of habit, it's dead. No ceremony.
3. Does it give me leverage? Not "is it neat." Does it let me or an agent do something we couldn't do before, so I'm freed up to go do the next thing? I keep what compounds my output. I ditch what's just a party trick, no matter how good the demo looked.
4. Value in the first hour. For a lean team, friction is the silent killer. If I can't get real value out of it in the first sitting, it's not for us right now. This is exactly where Buzz gets interesting, because the answer depends entirely on how you run it, and that decision is where most of the coverage falls apart.
Money leads. Leverage closes. Screenshot that and run it on the next launch, not just this one.
So how would I actually pilot Buzz without lighting a weekend on fire? Where's the trap-door most of the hype is walking you into? And what's the buried feature that turns part of your AI bill into hardware you already own?
Heads up, free readers: paid subscribers get the SaaSpocalypse Audit. It's the fill-in scorecard I run every tool through. You list your stack, score each one on the Four-Shoe Test, and the sheet totals up exactly what you're overpaying. Upgrade and it's yours.
