314xel.bsky.social
EU resident 🇪🇺
Slava Ukraini! 🇺🇦
Average coder. I'm also very modest. </>
Privacy advocate 🔒
Books! They magically make you less stupid! 📖✨
Save the climate, I need snow 🏂🏻
Live long and prosper 🖖🏼☮️⚛️
1,520 posts
912 followers
1,059 following
Regular Contributor
Active Commenter
comment in response to
post
Only upside I can see, it's Friday?
comment in response to
post
comment in response to
post
Ba - the "personality" part of the soul - "takes the form of a bird with a human head". The alternate hieroglyph also has a bowl with smoke coming out.
en.wikipedia.org/wiki/Ancient...
comment in response to
post
And his government made "arbitrary arrests and detentions targeting activists, journalists and others".
But he's articulate and speaks 2 languages, Kirundi and French.
Again, Burundi is ranked almost dead last by GDP per capita (185th).
comment in response to
post
Oh boy, my random choice led me to learn a bit about Burundi. So the president is Évariste Ndayishimiye, and there could be some parallels with Trump - he rejected Covid vaccines, and said that LGBT persons "should be stoned en masse".
comment in response to
post
Exactly. If this was the president of I don't know, Burundi (and I mean no disrespect to Burundi), I'd say "whatever". But no, this is the country with the largest GDP and nuclear arsenal. It's the worst, most unbelievable timeline.
comment in response to
post
He also likes to hear himself talk and give sermons on subjects he knows nothing about and that nobody asked for. Like a pretentious cult leader.
comment in response to
post
Despite Tesla limiting the Supervised Robotaxis to a small geofence, they also require the constant attention of a Tesla employee in the front passenger seat. The Tesla supervisors had their thumbs hovering over the door release button throughout the rides, which we suspect is a kill-switch.
comment in response to
post
Makes me wonder if the script was AI slop output.
comment in response to
post
She's an idiot, but the comments though... Is there anyone on Twitter not paying for the blue checkmark?
comment in response to
post
Not to be confused with "It's what God intended", that's an entirely different beast. And I don't recommend the appeal to the divine.
No no, this is pure folk wisdom. It's as if "to be, or not to be?" ever required an answer. Or if for some obscure reason you're not satisfied by "42" anymore.
comment in response to
post
Surge protection? No, more hoarding pls 😄
comment in response to
post
*sweating sheesh
comment in response to
post
I remember that I need an UPS whenever this happens and I go through "are my RPis SD cards fried this time?" while swetting a bit. Then I forget about it until the next powercut. I also don't have room for an UPS.
comment in response to
post
Which is fine, they just gave us the choice of tools, and I chose not to use it. Company still pays for a licence "seat" for me in case I'd ever change my mind. Efficiency 😂
And my coworkers are capable programmers not blindly trusting AI, still hurts to see them acting masochistic, though.
comment in response to
post
Oh jeez. If it would ever get to Cursor, I'd quit too. Fortunatelly, my employer isn't as insane and the work environment is otherwise healthy. The company did feel the pressure to hop on the FOMO bandwagon, and most of coworkers found it neat and beneficial. I'm in the *crazy* minority.
comment in response to
post
People have a natural bias towards trusting authority, and that includes machines, because historically, they were accurate instruments. This is the dangerous thing. It doesn't matter how good (gen-) "AI" will get, a 0.1% uncertainty will never help. It's a statistical model, but not used as such.
comment in response to
post
I don't care if hallucination rates fall to 0.1%. Sure, programming doesn't put people's lives at risk (most of the time). Point is about GPT in general. Should we then talk about, trusting GPT as a news source? Using it for medical advice? It's programmed to give you *an* answer, even if it's wrong
comment in response to
post
I know, and fully agree. The pressure at work was immense though, so I had to give it a try so that I wouldn't be talking out my ass and be a hater just for the sake of it.
comment in response to
post
I mean, it's amazing most of the time in "chat" mode for example to summarize a cron job file and explain what it does, what the possible parameters for the command are, etc (although it's not to be fully trusted, sometimes it hallucinates parameters 🙄). But in agent mode... no. Just no.
comment in response to
post
Climate change . #MemeForToday
comment in response to
post
Yeah, and streaming it on Netflix is only $5 / month, and you get to watch more than Stranger Things for the money. WTF? 😂
comment in response to
post
Even if it would get it wrong only 10% of the time, you still have to double check because of that, which means spending more time correcting instead of just doing the work.
I'm not even going to talk about it sucking the joy out of programming. Again, boomer talk, but I will die on this hill.
comment in response to
post
That being said, thing with coding and AI is, I'm not looking for perfection, I'm looking to minimize annoyances while doing it. And I feel that the Copilot agent is doing the opposite of making a programmer more productive. Top management has no idea about these details, it just followed the heard.
comment in response to
post
Thing is, I've seen the same thing happen to coworkers while they screen shared, reject 4 completely useless code completions in a row and finally writing by hand. Annoying to see. In their own words, they have a "love-hate relationship with Copilot", but still decide to use it.
comment in response to
post
I'm somewhat a perfectionist, that's true, and look to polish a feat as much as I can in the time given vs "just get the work done". I'm old school, and the business requires products out the door fast. If these become incompatible, I'll just look for another job instead of adapting and be unhappy.
comment in response to
post
That's exactly how I feel I'm perceived, a "boomer", and I'm 40. I have no problem with it. Coworkers can do what they want, if they find it helpful, it's fine. It's not my decision what tools they use, company pushes AI and bought licenses without asking, and the genie is out of the bottle anyway.
comment in response to
post
And time will tell if it will make my coworkers lazier and worse programmers, or more productive. I'll stick to StackOverflow when in need, and read all the comments not just the top answer.
Then again, about the distraction thing, I'm the kind of guy that rarely codes with music in the headphones.
comment in response to
post
I'm... vehemently against it at work too. Half my colleagues use Copilot at work. I don't. I tried it, and the code suggestions were the most distracting things ever in an editor. I hit escape in VSCode most times than tab, so uninstalled. Felt like coding with a non-programmer over my shoulder.
comment in response to
post
Hmm. I've been happy with these 2 feeds so far in staying up to date with the news:
Catch Up - 12h
bsky.app/profile/did:...
Catch Up - 3h
bsky.app/profile/did:...