I've had an idea for a different type of free & open search engine for the last 15 years. I ended up pursing other ideas instead over the years.
Last month however I decided to go back to the idea and give it a shot. Right now I'm in the process of scraping and building a huge index. The technical challenges have been plenty. But I should be ready to publish an alpha version by end of month or so.
It's been working for tens of thousands of years. What changed in the last few decades wasn't parenting or technology. It was the rise of the nanny state where the parents gave up the parenting of their kids and entrusted that to educational institutions instead.
Back in the day when I was like 15 and DC++ was still a thing, I used to browse people's shared folders. One day I came across a file called "the paradox of false positive". It was a 1 pager that described how a machine which is 99.9% accurate at identifying terrorists would be completely useless due to this false positive base rate fallacy you're describing.
It really stuck with me throughout the years. It's kind o remarkable how even a 99.9% accurate heuristic is insufficient at scale.
Which begs the question: lets assume the intentions are pure (which we know they're not but lets be generous), what other options are there when 99.9% heuristic is not good enough? how do you design systems when they're guaranteed to fail as they scale up?
Been using PHP for 20 years and have built multiple startups with it that went from ideas to exits.
Today's PHP is better than it has ever been. Are there some things that are rough around the edges? yes ofc. But there is no language that doesn't have that. It's all trade-offs.
Last week I switched from Nginx+FPM to FrankenPHP and my god even the deployment experience got 10x better.
Safe to say that if you haven't tried the language, give it a shot. Within a few days you'll know if it's a good fit for you or not!
I've seen sources like this before. It's all hearsay and promo. I was asking for any publicly available verifiable information regarding the cost of inference at scale. I haven't seen any such info personally which is why I asked.
I'm dying to see S-1 filing for Anthropic or OpenAI. I don't actually think inference is as cheap as people say if you consider the total cost (hardware, energy, capex, etc)
“people start with a k8s cluster for 5 users” doesn’t really exist
Most people in the BiP these days barely know how to deploy a database or host something using nginx. it's all vercel, supabase, aws, clerk, yada yada. Cost aside, I think that people are addicted to complexity.
I post some stuff on social media like code snippets and some markdown instructions. Most platforms have terrible support for rich formatting. So I decided to do it this way instead. I can now just post screenshots.
Also, when creating technical docs and presentations, this turned out to be super helpful.
Just a fun side project I figured others might find useful!
Last month however I decided to go back to the idea and give it a shot. Right now I'm in the process of scraping and building a huge index. The technical challenges have been plenty. But I should be ready to publish an alpha version by end of month or so.