Free
124 posts
Free
@aidrinksyoshake
"Execution is everything." Albert Pierrepoint
Katılım Nisan 2026
63 Takip Edilen10 Takipçiler
@mattpocockuk It's funny, because that's what human coders often do, unit tests just for the sake of it.
Imho the biggest value is in integration tests anyway (because most critical errors happen at the boundaries), but that's just me.
English
Adding a tweak to the /tdd skill:
"Do not add tests which simply restate the implementation. These provide zero confidence."
Getting sick of shit tests just to provide evidence of RGR.
English
I admit being a bad programmer, so maybe you are right, but even so I am curious what gives you the confidence that (apart from some niches) humans will still be needed to write and read code.
Because to me the trajectory seems clear.
And if a black box does what it was specified to do with a high degree of confidence, imho pure economics dictates not wasting ressources looking into it.
English
@aidrinksyoshake i understand how you can think that if you have never understood programming
English
@vlad_mihalcea I am convinced that in a few years reading code will be as uncommon as reading machine code is now.
Will be interesting to see what kind of tools will emerge for verification and observability.
English
Reviewing code generated by AI offsets all the productivity gains you get by using AI. It takes a lot of effort to review and validate changes, and you cannot do that for 8 hours per day.
Unless you are vibe coding, the actual productivity gains are marginal in large projects where software mistakes cost a lot of money.
English
Free retweetledi
bro it isn’t generally intelligent bro its only read every book and paper ever written and just making connections between them bro. its only thinking for twenty hours bro it’s just brute force thinking bro. its only solving erdos problems bro it could never be an accountant bro
English
It's because you havn't developed good enough software taste to see the subtle slop (this is not shade, I say the same thing about myself).
If it was obvious slop, like a 100 if-else branches to give you the first 100 Fibonacci numbers, it would raise the same alarms for us as text slop does.
This is not anti AI by the way, I think it's a modern marvel and that in a few years about the same amount of humans will look at code that now do with low level processor instructions instructions.
English
Wow, this is great, and without knowing much about physics, this makes sense to me, especially the entanglement explanation.
What I don't understand here though is that it supposedly follows that FTL communication is possible from the statistical deviations, would love to hear a more detailed (but still simplified ELI5) lecture about that conclusion.
English
I am optimistic that we will one day make contact to extraterrestrials because I don't think that the speed of light is a fundamental limit. Here I explain why and I have a brief summary below.
English
“AI will replace developers”
“AI will replace developers”
“AI will replace developers”
Cool.
Now ask AI to handle:
Kafka, Redis, Kubernetes, AWS,
circuit breakers, retries, idempotency,
cron jobs, containers, Docker, DLQs and many more.
Production will answer for us.
Don't fall for that shit.
English
@zuess05 Code stopped being the bottleneck for MVPs. Attention became the bottleneck.
English
@CACandChill I don't worry about finding another job, our core skill is problem solving, and that will always we valuable.
What I worry about is that there will not be enough people with jobs and disposable income left to pay for whatever that other job provides.
English
Software engineers, imagine AI becomes 10x better at coding in 3 years. what are you doing next?
English
Dario Amodei says no company has figured out how to make everyone a real participant in what AI is becoming
"something is happening to humanity at a scale unseen in centuries, and using the technology isn't enough"
Anthropic has tried, but there's still a missing piece
English
I think you are correct about verification (and I also think monitoring) being the holy grail, but I don't think it has to be mathematically (even probably can't be).
I think it will take a bit of time for "vibe engineering" to figure this out, and sadly suspect that there will be no silver bullet, like there wasn't in the past for so many things in software (it's always tradeoffs).
English
Hot take: AI code generation doesn't actually save you that much time.
If you have to painstakingly review and debug every line of AI-generated code, you're just trading writing time for reading time.
The real holy grail? Verification.
When AI can mathematically prove its code is 100% correct, you can confidently deploy it without ever looking at the source file.
English
@plainionist And I suspect that the first can quite often turn into the latter if the software grows (maybe even the other way round as well).
English
@petergostev I think it's a mixture of three things: trying regulatory capture, marketing, and genuine fascination and concern.
English
What I don't understand about Dario's constant scare stories about AI taking our jobs is what's exactly the goal here? To warn? Ok so what are politicians meant to do here, retrain everyone as electricians?
The exact effects of AI and the timings are so unpredictable that any planning is pretty much useless. It might well be the case that even with AGI there will be 5-10 years when there's little measurable impact on jobs - because of regulation, inertia and human adjustments.
The only effect we will get is people being upset and worried about AI, and wouldn't you be?
This is why you get people eating up nonsense about water use, or booing AI talks at universities and getting data centre build outs cancelled.
I genuinely don't get what the upside of this is, just to 'be honest'? Good impulse maybe, but he could also be wrong. To go out of your way and say this with such vigour seems strange.
Then the obvious answer is - how about not push the industry forward? What train Mythos? Opus 4.6 was already a leading model, there's no need to accelerate the 'devastating consequences' that he imagines.
Or is the goal to capture the lead, scare everyone and get all other labs banned? Can't imagine they'd believe this is achievable either.
English
@Kasparov63 He wants the stability that comes out of the barrel of a rifle, as they all do.
English
That dictatorships lie out of reflex never receives the treatment it deserves in the free world press.
Isaac Stone Fish@isaacstonefish
Stop saying that Xi wants stability. He doesn't want stability, he wants to increase China's power globally. That is intensely destabilizing. You can love China or hate it, you can feel like the Chinese Communist Party is good for the world or bad: stability in U.S.-China relations does not help Xi Jinping accomplish his international goals. Why do people keep making this mistake? Because they keep taking Xi Jinping's remarks at face value. It's a mistake journalists don't make nearly as often with politicians, but somehow, they decide that Xi communicates what he actually means.
English
Don't want to be a bad influence, but tbh I would just buy a new one (depending on how old it is and how much it cost), and tell myself that I wanted a new one anyway, because of features xyz.
And the newer ones also are supposed to be more repair friendly, like can't you just change the whole printhead part or smth (no idea, have my first 3d printer just for a couple of months) quite easily?
English
my 3d printer has been "dead" for like a month. stuck filamenet *somewhere* in the extruder or something. all procedures look complex af and don't wanna tinker with it. local services tell me to bring it to their place lmao if i ever unplug this thing and put it in a car
at this point i'll fly out someone to come and fix it and hang out for a few days 💀 any contenders
English
what are the biggest signs that you are a nietzsche reader?
English
@beffjezos I can only repeat the mantra: Hanlon's Razor.
No need to demonize the political opponent if the facts speak for themselves. I understand the emotions though, because the stakes are high.
English
They literally want to stop progress and widespread prosperity because if everyone is miserable they are easier to control.
Same reason why they love to have criminals everywhere and encourage stochastic terrorism
Payton Alexander@AlexanderPayton
Data centers are generating trillions of dollars in value for the economy. Data centers don’t consume water, they return it to the environment. Data centers lower household energy costs by producing their own electricity and generating stable demand that reduces operating costs for the grid. Data centers create jobs for construction workers, electricians, and other skilled workers year round, all across the country. Data centers data centers pay billions in state and local taxes without using public services, meaning more resources for the rest of us. Data centers power every part of the digital economy in ways most people don’t even know, even enabling things as mundane as stocking products at grocery stores. Data centers are driving the AI revolution that will keep America ahead of China and enable universal high income for everyone.
English