2,275 karma · joined March 26, 2015
Built Tunnelmole, the open source tunneling tool with similar capabilities to ngrok: https://tunnelmole.com
I'm also less afraid to make promising but risky changes. I have "broken" my machines multiple times, but each time this happens, I just boot back into the previous working iteration, with no real interruption to my work (or fun, depending on what that machine was for).
I can do things like try out new bleeding edge kernels and GPU drivers, important system libraries etc... getting the benefit of these things basically risk free.
There is also one server running NixOS serving up part of what runs Tunnelmole in production.
Initially LLMs were not good at writing Nix, but now they're much better and a good coding agent can do a write/build loop until their config works.
That is the road to hell. Even if your higher ups think it isn't, this strategy will for sure blow up in their faces soon.
I predict that Meta is soon going to be going on a hiring spree to fix all of the issues this strategy is going to create.
We saw this with the Bitcoin wallets recently, with hidden 5g modems being added in a supply chain attack. The payoff in this case for the attackers was large enough to motivate them go to such extreme lengths.
Interfering with a democratic countries elections is a pretty big incentive for across like Russia. They also have the resources to do it.
Given whom did this attack, I would not be surprised if it was used as intel for strikes on journalists.
There are reliable sources from inside the Israeli Intelligence apparatus that have stated on record that decisions were made to sacrifice 500+ innocent civilians to carry out a strike on one human target. I bet such a strike would be more likely if spyware indicated that one of the innocents was an "enemy" journalist.
It's way faster than Opus or any of the GPT models.
I have a coding harness which is opencode plus a few skills relevant to my workflow. Deepseek 4.1 Flash does very well in this environment. I haven't noticed much difference quality wise compared to Opus 5, which I use in my day job as my employer pays for it (although I'm considering using DeepSeek here too given how cheap it is).
I don't have RSI, but I do this whenever my arms get tired to make sure I don't end up with it.
For example, if you said to your counselor or psychologist that John Smith is going to rob the bank next Thursday, they will report that to prevent the imminent risk of serious harm to others.
And remember that anything but local AI you personally control is not anonymous. Not even AI you pool with your friends.
I find this approach is easier to debug and I can also use the tool myself to ensure it's working well.
Same with skills, symlink to skills at .claude/skills
Could be a strategy going on here.
Before they invaded Ukraine, Russia criticized Ukrainian military buildup as "destabalising". Probably with full knowledge of their own plans to invade.
In NZ, I approached the immigration desk with my American partner at the time, they stamped her passport with a visa but said to me I didn't need one.
Right. So expect thousands of Flock cameras to be hacked soon.
Possible perps include foreign intelligence agencies (Khamenei in Iran was tracked down for his assassination using Iranian traffic cams), stalkers, domestic violence perps tracking their victims, the list goes on....
More than likely though, multiple of the above.
No such feature in the other socials. Open the feed, you could see toxic content from an account you never followed thanks to the algorithm...
One AI generates code, another AI finds the security holes and systems everywhere get compromised.