imrozim

Wasn't talking about the client flag, agreed that's seperate headscale replaces the coordination server itself the part is tailscale inc actually runs and sees the connection through that's the server side pice

If metadata loging is a concern, worth mentioning you can self host the co ordinates server with the headscale instead. cuts tailscale is out of loop entirely rather than just opting out of what claim to collect.

Worth checking, elsewhere in this thread someone points out glm only assessed the damage after the fact, it didn't stop the attack, and Hf apprently never sought access to a trusted defender program with a closed model either the open model saved them farming might not hold up.

Any one knows how this holds up on long context retrieval (needle in haystack , ruler) vs same size full attention model? efficiency gains look great but that usually where linear attention hybrids fall apart.

Every node js project starts with npm install and suddenly you have a 500 packages you didnt ask for. Half of them haven't been touched in years.

Ai gives generic answer for ideas but it's great for code. Pattern matching works for one not the other.

I use ai to build a startup but I still decide what to build. Letting ai makes product decisions is where companies loose it.

Setting up AWS for my startup felt like a full time job gave up and just used a supabse. AWS complexity is real even for similar stuff.

I use Claude api for my startup and the billings and rate limiting hurt. But local models cant do what i needed yet. Wish they could.

I use postgresql for my startup but every time i needed a quick local testing i wish it was as simple as sqlite. No config just works.

Im 19 trying to break into teach and third exactly what scares me they killed the apprenticeship before i could get in how do you become Sara if there's no Ben left to learn from.

100% aggred, i learn coding by building stuff and breaking it when you let ai do everything you skip that pain and also skip the understanding.

Used to check every line for my project. Now i just check the tricky parts still don't know if that's ok or just lazy?

It's not ai that scary it's people using its field they don't know and then defending wrong outputs like they built it themselves

Fair didn't think about batching makes more sense for self hosted models then.

Yeah rules never work you just engineer around it I added a extra reviews steps on ai outputs because asking users to verify doesnt actually happen.

3x faster inference means cheaper api costs tooo. For solo dev building ai this matters a lot

Same experience here i use open router with claude as fallback for my startup. Is Kimi if close in quality the cost is difference hard to ignore

Good to know might actually try it for one project 1 st before switching

True but switching is not that easy when all your ci pipelines and integration on in GitHub.

As a solo dev GitHub going down is scary all my code, all my history, one platform. This makes me want to keep local backups more seriously.

A model from 1930 that thinks computer is a human job is wild we come so far 100 years

True but as a solo dev, you cant afford to pay for every tool you depend on. That's the trap.

This is scary as a solo dev who builts on postgresql. You pick a tool trust it, build around it, one day it stops. Oss sustainability is a real problem

Built a 4 startups and abandoned most every abandoned one thought me something I use today the graveyard is actually where my skills come from

This is why I never give a.i agents write acces the prod. Read only the most. The agent did exactly what it allowed to do

I use a.i to build my startup and it massively helps but i still spend hours reviewing and fixing what is genertes.

Forcing phone tab to add friend is genius kills boots and fake account instantly every social network dies due to fake connections.

PII masking is always an afterthought when building a backends 50m active params means it can actually run in production.