It's funny how over saturated that green is on my phone's OLED screen vs my desktop's LCD. The green is retro on my desktop my very not retro on my phone. It is supposed to be a subdued retro green not the overly saturated color that mobile makes it out to be!
The blue comes across truest to my vision on both LCD and OLED displays.
It depends how you use it. You can either get it to explain a concept, or do your homework for you. Its a bit like the decision students have to make as to whether to review their material before exams or go out partying.
Overall it just seems like a huge waste of money to piss away the huge tuition cost your parents probably paid.
You can use an llm to get out of doing homework but you can also use it to ask every question you would ever wanted in a 1-1 tutoring session. The problem is kids will use it to cheat on their homework. If we can’t deal with that problem then a ban is necessary. But these things can be phenomenal teachers if you use them properly.
As an educator, this is exactly what I struggle with. I'm pulling out all the stops to give students every chance to do the hard work and not lean on AI. But there's a good chunk of the class who don't listen to reason. I haven't figured it out yet. They know, logically, they can't pass an interview, but that's apparently a "tomorrow" problem.
The smart ones either use it not at all, or use it to positive effect, like you're saying.
> But there's a good chunk of the class who don't listen to reason. I haven't figured it out yet. They know, logically, they can't pass an interview, but that's apparently a "tomorrow" problem.
These people should be doing manual work, not intellectual work. There is no shortage of manual work available.
It's funny that GP mentioned science fiction as a negative because what immediately springs to mind, for me, is Neal Stephenson's The Diamond Age. We literally have the tools to build his "Young Lady's Illustrated Primer" today. We just have to give today's AI a lesson plan to follow and ensure that it never gives the student the answers, and only keeps explaining the concepts in different ways until they click. Wrap that in an iPad app and you've essentially got the exact self-paced learning tool that Stephenson envisioned changing the world.
And how do you propose that to work if the internet is still full of AI services that just give you the answer or write your essay? The only way an Illustrated Primer can work if you can’t trivially cheat. Which is to say, it solves nothing compared to the current situation.
Why are you assuming that all students will cheat? That's entirely unreasonable.
But if you want a way to push cheaters towards doing the work then a boarding school without cell service would work. Although I guess these days you'd also have to account for starlink and the like.
Probably easier to warn everyone that they will almost certainly fail if they cheat and then be entirely uncompromising when it comes time for exams. Not your problem if they knowingly shoot themselves in the foot.
Public education is about forcing kids (the vast majority of whom are literally incapable of long term planning) to become educated so as to have a stabilizing effect on society.
An uneducated populace is easily lead astray, am existential threat to a democratic society.
They are great for self-teaching and great to cheat and not learn anything, depending on how you use them.
Main problem is that the technology was very disruptive for education and nobody has figured out yet how to utilize it at scale for schools and universities.
Why not though? I'm pretty sure many (especially this crowd) have childhood memories of playing with calculators before really knowing how to add.
A ton of natural questions arose like how it differs from tallying or concatenation, what the decimal point means, what negative means, ways to get to zero, ways to overflow the display, what even is overflow, what the other buttons do, etc.
The limitations are very obvious very quickly and their frustrations only further motivate the student to learn it properly.
I think beyond all else, what you're really proving is that this is more political ideology than a genuine concern.
I'm in Seattle and everyone knows exactly where human trafficking is happening and the police are doing nothing about it. Teenagers are being pimped out all along Aurora and literally nothing is happening despite literally years of public outcry.
The pimps get arrested again and again and then released without charges being filed.
My team (Microsoft Band) discovered the reason why the surface's keyboard sometimes wouldn't work when connected. There was a hardware bug in the cortex MCU the keyboard used involving waking from deep sleep. One of our FW engineers spent several months figuring it out and eventually reported it to the manufacturer, and to the Surface team. IIRC it was something about wake on interrupt in a specific deep sleep mode and also something around timing.
It was a rather nasty bug. Firmware is full of nightmare scenarios like that.
You clearly have never tried to implement sleep on a microcontroller. It has nothing at all to do with ACPI. And this kind of eldritch bug is par for the course. It has nothing at all to do with Microsoft or PCs at all. Microcontroller sleep just sucks in a lot of incredibly weird ways.
On Band we had one of the top firmware engineers in the world who got our power situation setup and figured out that bug.
We had another bug involving a single cycle timing tolerance of memory on a bus, we were one cycle off in our config. We dual sourced memory chips and one of our suppliers had chips that worked with 1 fewer wait states, the other chip didn't. Once we started blending the 2nd chip into our production line we had weird issues.
One of my guys spent ~2 months thinking it was a software bug. We had a null ptr deref in our code, he went line by line and by the end of 2 months we had literally bug free calendar code.
Turns out the timing bug manifested in the exact same bit being corrupted every now and then in memory, which happened to be where our pointer lived at. We figured it out when a new production build caused memory to be moved around enough that someone else's pointer was getting corrupted. :)
So then the HW investigation began.
The principal engineer on Band's firmware refused to allow us to have any open crash reports, all crash dumps had to be fully investigated, root caused, and fixed. In a few cases we found customers who had faulty hardware and we reached out to individuals to replace their units, and brought the broken units back to our labs to figure out what went wrong on the assembly line. (Although on occasion stuff just gets assembled wrong, even with slews of automated tests on every unit that leaves the factory floor!)
That same principal engineer reviewed every single line of code that went into the Band's firmware. That code base had a consistency of code, style, and thoughtfulness that I'll likely never see again in my career.
> The trouble I find with LLM generated recipes is they miss the nuance of the technique. Often the success of a depends on a single step or ratio. For instance “fried chicken” has a million incarnations the world over, but you can’t just average out the recipes and end up with tasty fried chicken.
Specify what technique you want. Explicitly say you want to correctly follow all the techniques of the chosen cuisine.
All the LLMs have ingested nearly every cookbook ever made, across multiple languages.
You can upload a photo of your spice rack (with visible labels) to ChatGPT and tell it to save your pantry ingredients as a memory.
LLMs are absurdly overpowered for cooking, when used right. If you ask it for a week long meal prep plan the results will be meh, but ask it for kheer inspired rice crispy treats (which everyone reading this should to, kheer rice crispies are the best!) and you'll get some solid results.
You may notice at first the LLM will still water things down for "American" tastes. With Claude/ChatGPT you only need to remind it once or twice not to do that and it'll course correct all future conversations.
> All the LLMs have ingested nearly every cookbook ever made, across multiple languages.
That's not a positive thing, good recipe developers are Rare. For every recipe that's been meticulously tested and documented there are 1000 that haven't been. Many cookbooks are riddled with errors.
Sure, but most recipe books are just copies of other good recipe books. There are only so many ways to bake cookies.
I've always been a pretty good cook, but I've been able to pull off some really cool stuff with the help of ChatGPT lately. It is probably just an incremental lift, and I still catch it making errors from time to time, but it has been a huge help in the kitchen.
reply