In college I needed a math class, so I found one and signed up. The first day , as all the students(except me) whipped out their "sliderules", I realized I was in an engineering class...lol
I gave Copilot a shot. It seemed like everything I asked for was something it wasn’t allowed to do for some reason or other. Not sure if that’s a comment on it or me, but I uninstalled and remain unimpressed.
Here’s a simple example of how letting technology do the hard work has made us lazy: how many people’s telephone numbers could you recall without having to look them up on your phone?
Back in the day I had memorized the digits of all my frequently called family members, friends, acquaintances and businesses. Now I worry that if I lose my iPhone, I’ll be in big trouble until I can get a new one and reconnect to the cloud.
Oh my God, YES. I spent a bunch of the 1980's calling BBS's with my computer, and I literally had about 100 numbers committed to memory I dialed them so often.
Why aren’t babies born walking if that’s the way humans end up anyway?
Without a solid foundation, like getting from point A to point B and so on, one may as well build that proverbial house on sand. There’s nothing like bedrock to build anything on and I fear that, while convenient, chat gpt, copilot or anything along those lines will produce results built on thin air. There’s a saying I remind my old self a lot these days, “Just because you can doesn’t mean you should”. The wizards of tomorrow might do well to be reminded of that occasionally.
Funny you should mention map reading, as that’s exactly what I was thinking of. And yes, I think that’s what will happen. I think understanding is as important as the answer. And that the struggle to understand difficult concepts is key for development of new ideas. Last time I had physics or chemistry was the late 70s and we weren’t steadying Hamiltonian physics or quantum mechanics (though I did read The Dancing Wu Li Masters back in the day 😂). No calculators throughout my entire schooling, including SATs. But we learned lots of math “tricks” and I, too, knew how to use a slide rule!
To be honest, it is about the third year of your physics degree when you get introduced to the Hamiltonian, and I can assure you that it BLEW my mind when I did learn about it.
Like Keanu Reeves in Bill and Ted's Excellent Adventure blown away.
I am old school. Graduated high school 1970. Did some key punching for my friends because I could type but you are way over my head when it comes physics and theories. When I went to college after the service I tried majoring in coffee shop and pinochle. I like a pencil because it has an eraser.
I too am old school. I don't mind the fancy new tools, but if you don't understand the foundations, then you never will know if it is giving you good answers, or if it is full of shit.
I’m of the old generation where my dad showed how to use a slide rule in Grade 8. Calculators came into the class soon after and made the slipstick obsolete. But it did teach me to do fast order of magnitude calculations in my head. Even today I can do sums, multiplications and percentages before my colleagues whip out their phones and calculator apps. Quite often I can spot arithmetic errors in financials or contracts. I worry that with AI, especially engineers and scientists will lose the skill of seeing through the numbers. It does not take much to be slightly off and lead to an accident…
Yes, I learned to use a slide rule in high school, and to this day, I impress my colleagues when I can estimate pretty darned close to the actual result (I also keep an old HP 15C on my desk, and it does get used at least once a week).
My biggest fear is that people will just rely on what the Chatbot spits out and never learn enough to know when it is hallucinating. I already see that happening in my current work.
And I worry that executives will soon replace plenty of people with GenAI even though it isn't really good at "novel thinking" purely to save money.
It may be great for overviews, but try getting into the details with it and you’ll soon see the problem.
Absolutely agree. But it does enable people to get an "answer" but no understanding of how you get the right answer, and why it is right.
To me, doing the work is key to actually learning.
I’m recommending that educators put questions to students that *require* the more detailed answers that AI is bad at.
When I was in uni, I studied physics, and if you wrote down the answer without showing your work, you got a 0.
It is all about showing the work. (Of course, this was in the mid 1980’s, computers were not common
Yeah, what you said before…
In college I needed a math class, so I found one and signed up. The first day , as all the students(except me) whipped out their "sliderules", I realized I was in an engineering class...lol
I gave Copilot a shot. It seemed like everything I asked for was something it wasn’t allowed to do for some reason or other. Not sure if that’s a comment on it or me, but I uninstalled and remain unimpressed.
I wrote a thing about AI a while back:
https://substack.com/@mommadillo/note/c-60951455?r=iggr&utm_medium=ios&utm_source=notes-share-action
Here’s a simple example of how letting technology do the hard work has made us lazy: how many people’s telephone numbers could you recall without having to look them up on your phone?
Back in the day I had memorized the digits of all my frequently called family members, friends, acquaintances and businesses. Now I worry that if I lose my iPhone, I’ll be in big trouble until I can get a new one and reconnect to the cloud.
I used to BE a database. Now I just ACCESS one.
Oh my God, YES. I spent a bunch of the 1980's calling BBS's with my computer, and I literally had about 100 numbers committed to memory I dialed them so often.
Why aren’t babies born walking if that’s the way humans end up anyway?
Without a solid foundation, like getting from point A to point B and so on, one may as well build that proverbial house on sand. There’s nothing like bedrock to build anything on and I fear that, while convenient, chat gpt, copilot or anything along those lines will produce results built on thin air. There’s a saying I remind my old self a lot these days, “Just because you can doesn’t mean you should”. The wizards of tomorrow might do well to be reminded of that occasionally.
Funny you should mention map reading, as that’s exactly what I was thinking of. And yes, I think that’s what will happen. I think understanding is as important as the answer. And that the struggle to understand difficult concepts is key for development of new ideas. Last time I had physics or chemistry was the late 70s and we weren’t steadying Hamiltonian physics or quantum mechanics (though I did read The Dancing Wu Li Masters back in the day 😂). No calculators throughout my entire schooling, including SATs. But we learned lots of math “tricks” and I, too, knew how to use a slide rule!
To be honest, it is about the third year of your physics degree when you get introduced to the Hamiltonian, and I can assure you that it BLEW my mind when I did learn about it.
Like Keanu Reeves in Bill and Ted's Excellent Adventure blown away.
I am old school. Graduated high school 1970. Did some key punching for my friends because I could type but you are way over my head when it comes physics and theories. When I went to college after the service I tried majoring in coffee shop and pinochle. I like a pencil because it has an eraser.
Agree. Hard to learn if you don’t have a basic foundation. I think of Admiral Rickover and what he to learn about nuclear propulsion.
I worry about AI and algorithms because like other things in our world both can be manipulated.
Feel more secure that they are people like you who are investigating these sciences. Thanks.
I too am old school. I don't mind the fancy new tools, but if you don't understand the foundations, then you never will know if it is giving you good answers, or if it is full of shit.
I’m of the old generation where my dad showed how to use a slide rule in Grade 8. Calculators came into the class soon after and made the slipstick obsolete. But it did teach me to do fast order of magnitude calculations in my head. Even today I can do sums, multiplications and percentages before my colleagues whip out their phones and calculator apps. Quite often I can spot arithmetic errors in financials or contracts. I worry that with AI, especially engineers and scientists will lose the skill of seeing through the numbers. It does not take much to be slightly off and lead to an accident…
Yes, I learned to use a slide rule in high school, and to this day, I impress my colleagues when I can estimate pretty darned close to the actual result (I also keep an old HP 15C on my desk, and it does get used at least once a week).
My biggest fear is that people will just rely on what the Chatbot spits out and never learn enough to know when it is hallucinating. I already see that happening in my current work.
And I worry that executives will soon replace plenty of people with GenAI even though it isn't really good at "novel thinking" purely to save money.
Thanks for your comment!