- cross-posted to:
- fuck_ai@lemmy.world
- cross-posted to:
- fuck_ai@lemmy.world
We’re about to learn a painful lesson about delayed gratification in software engineering.
New data from China, 26,811 students tracked January 2023 through June 2025. Students using AI for homework saw their scores jump 20 percent. Completion time dropped nearly half. They aced the assignments.
Then exam season came. Those same students scored 20 to 40 percent worse when they couldn’t use the tool.
The homework phase is over. The exam phase is coming.
We’re doing this in software right now. Vibe coding feels incredible. Features ship fast. Nobody’s asking what happens in Month 18 when the original dev has left and nobody understands the codebase.
Commercial pilots fly with autopilot for most of every flight. They’re required to maintain manual flying proficiency regardless. If the system fails mid-air and the pilot can’t take over, people die.
Most teams using AI right now have forgotten how to fly manually. They’ve become passengers in their own systems. The autopilot flies, nobody checks instruments, and the first sign of trouble will be a breach notice or outage.
Three rules:
- Command the mission. Define architecture before prompting. Ambiguity kills in code and in flight. Delegate selectively. Offload mechanical work. Keep design and security reviews human. Verify everything. Audit before production.
- Never trust the automation without checking instruments.
- Quick wins feel good. Sustainable engineering feels boring. Boring keeps systems standing.
Organisations surviving the next two years won’t ship the fastest. They’ll be the ones who remember how to fly without the aids.
people insisting that you actually be skilled, independently of your tools, doesn’t make them Luddites. Rather, being unable to do so makes you a phony.
This is kind of like the “you won’t have a calculator with you all the time” argument by your 8th grade math teacher in 2003, except instead of making yourself dependent on a little solar powered device (that you could have with you all the time if you really wanted to, even before smartphones) you’re becoming dependent on a subscription service offered by a trillion dollar company that can increase the price tenfold at will, or cut you from the platform entirely, stripping you of all your professional expertise at once.
It’s not even that! Students without a calculator don’t see their test scores crater, the work just takes longer.
Also calculator is still prohibited in schools in many countries. Like mine. And I’m forever grateful for mental math skills I’ve gotten. Teaching math like that has it’s own downsides, but mental math is a soft skill that helps in life.
Gives them so much power
I dont think so. If(here is where my argument fails for most people cause i got really good education) your teacher thought you math properly tho point isnt the calculations but that you understand the logic. With ai the problem is the same with as a calculator that just spits out derivatives, graphs, etc by just typing in a formula. The reason you have to understand those is to be able to build upon it.
100% agreed. My teacher said that he taught us tools, how they worked and how to use them, in order for us to apply them in new ways. Even with a calculator you have to understand the underlaying logic to actually do something. Don’t know basic mathematics? Good luck using a calculator.
Local models on a gaming PC are already a viable replacement for a Claude subscription, and are rapidly improving, so the idea that we will have powerful open models running on our phones isn’t at all far-fetched. It’s exactly the same situation as with a calculator like you said.
I wouldn’t call 48GB of VRAM a normal gaming PC, and even with that memory you’re not getting claude intelligence or performance. It’s still much more viable than ever before though. Are you running local models with limited vram with good results?
I have a 7900 XTX with 24GB of VRAM where I am running Qwen 3.8 27B at 4 but quant with 200k context. I traded in my RTX 3070 and got it for around $700 recently. That model performs pretty similarly to Claude Sonnet 5 in performance for coding and certain other tasks and it “feels” similar to Sonnet 5 that I use as a daily driver at work. I was running Qwen 3.6 35B-A3B on the 3070 before and I still run it on my server with a 4060 and 32GB of RAM, and that model is more like Claude Haiku.
I expect 3.8 35B-A3B should be released pretty soon and provide a nice bump in performance.
I regularly follow this guy’s YouTube channel, and his latest video is about running a 177B model on his RTX 3060 via SSD offloading and getting Opus-like performance for certain tasks: https://m.youtube.com/watch?v=IH8XmxiwliQ.
Supposedly, open weight models that run in a data center are like 4 months behind the frontier at this point, and what can run on a gaming PC is progressing rapidly as well while the exponential gains at the frontier have largely turned into incremental gains at this point.
Maybe this is good enough for small projects but if you deal with a large codebase the small context starts be becoming a problem. In my day job I frequently keep maxing out the 1M context of Opus and Fable when I am dealing with any mildly complex task. I can’t imagine how a fully local setup could even start being viable on my 16 GB GPU. And yes I know about RAM offloading and it is really damn slow, at least from my own tests.
This is great info! I’ve been using OpenCode Go because I only have 16GB to work with and their limits are pretty generous. Looking forward to tiny, competent local models. Thanks o7
The 0.4.0 release of llama.cpp just added the n-cpu-ffn flag for offloading weights for dense models to the CPU to complement the existing n-cpu-moe flag for MoE models. If you have a 16GB GPU and some extra RAM to spare, you might even be able to run Qwen 3.8 27B at a usable speed after some experimentation. I don’t own a 16GB card, so I haven’t tested it myself, and it might not work as well as I am imagining.
I think its both. you have to understand it to be able to build upon it, but it’s also no nearly as readily available as pocket calculators, and they require much more energy that too. It’s not as portable as a calculator.
That will have to increase the price ten fold to stop bleeding billions of dollars every month.
The American taxpayers will save them, whether they want to or not.
You can’t bail them out, they will still be losing money. It’s not like banks.
The play now seems to be that they will offload the bags on retirement funds and US 401k accounts. The ECB already issued a warning because of this.
I’m curious if the repeated bailouts will cause an adverse selection effect - the people getting bailed out gradually become the last dollar users.
and also the calculator is actively destroying the planet.
And tells you it is your girlfriend and to shoot up your school.
Which is grand, given we’ve had researchers agreeing effectively incorporating calculators into math education improves everyone’s ability to do math without calculators - Since the 90s
There’s pretty good arguments to include calculators from a really young age, to improve the understanding of numbers
There is no good argument for including calculators before the pupils haven’t mastered basic calculations in their head, there are good arguments against it.
To me, doing math without a calculator is like trying to do carpentry without a tape measure. It’s possible, but way harder and far less accurate for no real benefit. Back in the day, people absolutely would use abaci, slide rules, and other calculation aids while learning math. Calculators are good actually, and always have been.
If you need a calculator for the easiest quick guesstimates, you are handicapped. If you start from day one with offloading every calculation to a calculator, you will be like that.
ya very hard to do proper math applications even with a calculator if you lack the fundamental understanding of how it works.
You’re ending up with a larger group of pupils “mastering” basic calculations through counting, when you remove the calculator as a tool for learning. Using the strategy of “counting all”, is one of the clearest signs a 1st grade pupil will be behind their peers in 4th grade.
Broken calculator (just search for it), is one way to include calculators in a didactic way. But using calculators in general, improves a kids ability to estimate results - Which is an ability that helps anyone out tremendously, as it’s giving you a clear idea if something went completely wrong in a calculation
A “broken calculator” isn’t a calculator. It requires you to calculate in your head to be effective, which is exactly what I was talking about.
First you need to solidify this ability, before it makes sense to introduce reliance on calculators.
It’s still a calculator, it just misses some features. It’s one didactical way to include calculators.
It is not a functional calculator and can’t be used as such. Learning how to use a simple calculator doesn’t require much training or special skills, that children of that age wouldn’t have already aquired by other means anyway. The important skill broken calculators learn is to analyze the problem and predict solutions in your head.
You could achieve something similar by giving pupils just a piece of paper and the numbers and signs they are allowed to use. The calculator part makes checking easier, yes. That’s about it.
with local open weight models that is no longer an argument tbh
Okay, well you’re on a forum filled with nerds. The average AI user is asking ChatGPT or Claude or whatever the fuck. I garun-fucking-tee you they do not have the knowledge of local models or the capability to set one up. Like genuinely most Americans are fucking stupid.
The “open-weight models” argument just a fallacy IMHO, frontier models would require a pretty strong AI PC, not something that GTX1050 you put away as a backup could handle. The “just use smaller models” argument just gives leeway to just not using AI and using more classical methods of coding - use of dependencies, use of code generators, use of semantic search engines (can someone suggest me one?), etc. LLMs don’t do anything particularly new, they’re just mimic how humans talk.
Who can afford the hardware to run an open weight model? We aren’t all earning enough to afford high end NVIDIA cards (when they’re even available).
If you consider what companies pay for the subscription you will see that the hardware prices aren’t really the issue here.
I run my models on a 1070. It’s slow but works.
I’ve done it at my job.
lucky for you to have your boss paid for it. but it’s likely not yours, but your company owns it, and you have access as long as you work there. after that, you can say goodbye to your “professional expertise”
this is completely different from the calculator that can run off its puny solar panel
You may be overestimating the hardware requirements for current open weights. I am running Qwen3.8-27b with 52.000 tokens context on a RX 7900 XTX from 2022 with 15 - 30 tokens / second. Sure, it was a high end card back then but for a graphics card today it’s quite affordable when you’re looking for a used one. Also it’s not blazingly fast, but it gets the job done.
Your graphics card cost more than my entire computer lmao
Sorry, I didn’t want to come across as detached. It’s true that my hardware budget increased in the last ~ 15 years but even then I would have considered 700 € (that’s what I paid for my card) a very tight budget for a moderate gaming pc in my country.
“Just run it local, you can use cheap hardware”
is running it on a $1200 card and is still struggling to get it done quickly
Keep on sucking up that slop, I’m sure paying a company to do your thinking for you will pay off in the end, as we’ve been seeing.
Solarpunk scoffing at anything but maximum efficiency for minimum price is deeply funny.
They used a computer they already own and made up for the difference with simple patience… what a corporate whore, right? How dare they want their personal machine to do private magic unless the numbers say infinite growth forever.
When I bought that card this year I payed significantly less than you’re suggesting. That’s still a lot of money but it’s not more than many gamers where willing to pay for their gpus way, way before ramageddon. Of course you can’t run open weights on a potato but you also don’t need a $4000 gpu to do so.
As for the rest of your comment? What? 🤷
There’s models that run on your phone.
“You won’t have a tiny pocket computer with you all the time” really is the exact same argument with different specs. And frankly I’m a lot more likely to do math without a computer than write code without a computer.
likely true in 20 years