Recommend the Mom Test as good place to start on how you can make sales before having a product (and figuring out if your product is something people actually want):
Kinda? One of the biggest advantages of being small is that, when someone asks where they can find suchandsuch feature, you can build it and point them to it in the same time it would take a larger company to reply “yeah that’s on our roadmap”. You’ve gotta leverage that to compete.
There's always been vendor lock-in. Someone has to implement technologies to the standards, those are most of the time commercial interests.
Apache vs nginx, rhel vs debian, spring vs j2ee, etc. etc.
I think it's good for the young crowd to get into self-hosting. You learn more fundamentals that way, for sure. But I'm not looking to move stuff into self-hosted.
I get unlimited scalability in AWS. I can standup and run an app for $0-$5/month depending on how I want to build it. That app will have unlimited scalability, redundancy, automated backups, etc. It's not worth dealing with the headaches of managing things myself.
Most everything is the way it is today for good reason, it's just that the youngsters haven't had to learn those reasons themselves.
We don't need to fix hosting. There's no such thing as "portable" or "generic" and there never will be. At a certain point you have to use an implementation of <thing> and that's going to lock you in.
Like as a deployed component in our customer facing app stack? Not being used in that context that I'm aware of. I think there's a couple use cases being worked on though.
Are devs/engineers using MCPs to do their work? Yeah. Datadog MCP, Atlassian MCP, Gitlab MCP, backstage mcp, and we have our documentation behind an MCP. All used daily.
In my professional life I've removed everything besides MCPs that connect to our knowledge base, gitlab, and jira.
For personal projects I never used them.
It was never that hard to type a couple extra sentences and it always felt like people were using an LLM to create most of these skills/plugins... and if the LLM can generate them already then what good is it to keep more garbage in context?
I like it. I built a chrome plugin for rating the best dish at a restaurant, this definitely reminds me of that.
Added a couple spots in Seattle (where it's exceptionally hard to get any food below $10)
No suggestions, just happy to see something that's NOT an AI tool. Even happier that I'll get some use out of this. Maybe add somewhere on the site/app where I can input suggestions if I were to have any later?
Most of the AI tools people are building have little to zero value.
If you're making valuable tools, anthropic, openai, or one of the other big players in the space will hire you/buy you out and then they will release the tool themselves.
I'm getting really tired of everyone doing their own special rain dances and then acting like they made it rain.
100% agree. AI isn't hacking, it's a completely different crowd. So much front page space is wasted talking about the latest version, some funding controversy, or these useless add-on tools.
Can't tell you how much garbage AI tooling is being built within my org. Whole teams spinning their wheels basically rebuilding claude code because execs are obsessed with "dark factory". Individuals building plugins/skills/mcps that are duplicated or just plain wrong (why do I see so many skills/plugins telling our agents to use main branch when we've always used master???).
I've yanked MOST tools OUT of my claude code/opencode/codex. MCPs to centralize auth to our knowledge base, jira, gitlab. Then a singular short simple markdown file in each project/repo. When brainstorming, maybe I'll create a one off .md that I'll delete once the thing is built/implemented.
No, I think there is. Too many people building AI tools. I see it in my org. Tons of low value work but the ai sycophant is telling everyone "you're a genius, this is great" so they continue.
We built a "fire and forget" agent that's supposed to complete whole jira tasks or that we're supposed to be able to call from slack. You put a ticket in to "make random.md document" and it creates a ton of unit tests and makes "notrandom.md" and says "success". Tons of engineers pulling plugins/skills/etc and injecting them in places they don't belong, lots with blatantly wrong info. People making whole frameworks and tools around their AI workflows, but they're worthless, anything that made sense gets released a week after they finish building from anthropic/openai/etc.
reply