Don’t let media of any kind (news, social, etc) distract you. They thrive on your attention and in the end you’ll have many hours/days/weeks of frustration and resentment to show for it.
What has become possible with AI these days is nothing short of insane and now is the time to build. That nagging problem you’ve had with that product or service the last few years? Replace it.
In the last two weeks I’ve stopped listening to any podcasts, browsing hacker news, and anything similar and in that time I’ve built two separate production grade systems to solve real problems I had both at work and personal life.
One of them replaces software I used with family for years and the other was something entirely new that saves me a lot of headaches at work. Minimal dependencies and none on third party platforms.
And although I’ve been writing software for two decades I built these without writing a single line of the code. Full test suite, manually reviewed, etc but the point is that we’re at a point where almost anyone can do it. Most of that was written while I was doing other things like eating lunch, in a meeting, or cooking. Don’t wait for someone else to build your app for you.
Bitcoin itself is a filter: it’s an economic filter. Fiat is no filter.
Filters are a form of firewall.
I can’t shake this idea of using bitcoin as part of a network firewall. Anyone know of good papers on this topic?
#asknostr
If there exists such a technology that would give you total dominance over a critical domain and you can’t guarantee you’ll have a monopoly on such tech then your best move is to work on preventing anyone else from getting that monopoly. That is the core principle behind decentralization.
LLMs as an ad and modal blocker?
You could have an LLM (running locally ideally) that reads the content on the page and automatically closes the modals or deletes the ads. Decent at content moderation.
The bad news is this opens a can of worms. You’re letting a model that you didn’t train (more than just fine tuning) filter and manipulated the content you see and therefore trust the trainer to be aligned with you and your interests.
Meta and friends have surely figured this out a long time ago and probably why they release open weight models in the first place. It’s an interesting idea though.
