Ben Werdmuller has a good post today about returning to the distributed publishing roots of the web and thinking about how technology should redistribute wealth and power to many people:
It starts with software designed for people rather than for capital. The web once thrived on protocols instead of platforms — email, RSS, blogs, personal websites — before closed networks turned users into data sources. We are now seeing efforts to return to that ethos.
There is a lot that I love in Ben’s post. I also think it captures some frustration toward AI from some open web proponents. That’s not an opinion I agree with, though. For example, this part in Ben’s post:
Even the productivity gains that are being realized through use of AI tools are benefiting a small number of wealthy companies rather than individuals. This is the exact opposite of the power redistribution that led to so many people seeing such promise in the web.
That doesn’t ring true to me. I expect AI is benefiting a lot of tiny companies of only one or two people who are hardly wealthy, maybe even barely profitable.
AI does have the potential for harm. Let’s not gloss over that. But at its best AI can improve education, making all the world’s knowledge more accessible to more people. It could help people who aren’t fluent in English communicate better with their peers across the world. I don’t think the open web and AI are at odds.
Bob Booten AI should be opt in not forced into everything whether we want it or not.
Bob Booten AI should be opt in not forced into everything whether we want it or not.
Jamie Thingelstad YES for sure! We can even like crypto and the open web, as I do. Can I like document databases and the web? Yep. Can I like cloud solutions and the open web. Yep. Can I like Apple and the open web? Surprisingly yes. 👍
Jerome (He/Him) @manton I’d argue the wealthiest (and by reason, largest) of companies are the ones most struggling to reap the productivity gains of AI right now.
As a tiny company, we don’t have to worry too much about vendor control and we’re often encouraged to make our own calls on tools.
Manton Reece @bobooten.bsky.social I think opt out is fine, but I agree users should have control over whether it’s used.
rom were you talking about AI or GenAI?
JL Gatewood I like em all. Technology is fun and interesting so why not? But we should remember they are tools to be respected and understood.
I also like to put limits on what does what, and self-host wherever possible— for security, sure but surprisingly it’s easier for me to learn this way too.
Manton Reece @rom Generative AI. I’ve usually been shortening it to just AI now.
rom well, that’s the thing with making GenAI = AI when AI > GenAI — reduces other machine learning efforts and bunches it together with the “bullshitting” models. I know you know this, but hope you can help educate the others, too.
Mandaris Moore @rom in light of all the ways that technology has let me down, I'm kind of glad that we can reach out to different companies to do different things. It's more fun this way.