You hit the nail on the head here about how the corporate personalities are really starting to show themselves and reveal the desperate profit motivations behind the touted AI βeducationalβ benefits on offer. I do think we will look back on this moment perhaps not with incredulity (mistakes with social media and kids are now obvious in retrospect) but something like criminal negligence that we watched the same thing happen again. The Economistβs leader this week is on how AI is rewiring childhood.
My hopeful moments these days are based on a sense that the executives at the tech giants are as confused and worried as the rest of us. I am sure there are some very rich sociopaths in Silicon Valley, but the majority of the managers and engineers and marketers are human beings who want a better world for themselves and their kids.
I think that's true. Hopefully some are paying attention to what's happening in their children's schools. My optimism for how AI can be helpful in the learning process diminishes by the day, not because I don't think there is enormous potential - the things AI has enabled me to do are extraordinary - but because the scope of the problem seems to be so overwhelming and even worse beneath the surface that I just don't think school leaders truly grasp the magnitude of what's already here, let alone what's on the horizon.
I would like a Picard Mode in my AI models so that they would address me as the captain explicitly as a computer, without all the frippery or any "would you like something else, sir."
I like it. One strategy for disanthropomorphizing chat is to make the role play more specific. I have a friend who uses a robot voice any time he talks with Chat. It helps those around him understand he is addressing a computer and not them. I suspect it also helps him remember that what he is doing is quite different from talking to a human.
There is a well-documented history of Silicon Valley CEOs restricting their kids' technology usage. Despite driving the digital age, leaders like Gates and Jobs limited their childrenβs access, and I suspect current leadership does the same. A Silicon Valley paradox? Unfortunately, not every family has the time or money to establish these boundaries or resist the marketing that suggests children need these devices for educational purposes.
The consequences are significant. Devices are not substitutes for parent involvement, struggling to learn something new and challenging, or even children playing together. The decline of children playing outside erodes the mental and physical resilience that comes from real-world interaction.
The tech industry is now making AI the next frontier, which risks creating a generation unable to learn the fundamentals of their chosen fields. You know by working to understand concepts and solving problems. If AI provides easy answers, many will take the path of least resistance and enter the real world dependent on the software. If the AI isn't available, they will struggle, much like those who cannot perform arithmetic without a calculator. However, the risk here is significantly higher than in previous technological shifts. While a calculator might replace mental arithmetic, AI attempts to replace a broad spectrum of cognitive abilitiesβfrom writing and solving problems to coding and critical analysis.
I am not against AI, but it must be used correctly: apply yourself, struggle to understand what works, and use the tool as a tutor rather than an answer provider. Spending $100k to $400k in college on a college education is a waste of 18 years if you don't know the basics or do not know how to write, analyze, and solve simple problems. If the goal is merely a degree without learning, we should not be surprised when graduates struggle to find employment. The technology is here to stay, so we must teach future generations to use it for productivity, rather than letting it think for them.
Parents also play a significant role in ensuring that children are not focused solely on a piece of paper or specific grades, but instead put themselves through the intellectual struggle that defines real education. This is how I was raised, and it turned out fineβthough I admit it is easier said than done now than it was a few decades ago.
βintellectual struggle that defines real educationβ well said. The deeper the knowledge, the more polymathic cross referencing possiblities arise to meet awareness. Every βhowβ needs a βwhyβ. To choose why we outsource learning to leaning on AI would be to separate flatlined human thinking, quick answers to questions, to more deeply answered dilemmas with steps of probity that include inferences from a lived life of experiences. Holistic thinking is lived learning complexity, not just programmed computing. Perhaps we will get there as in William Jamesβ βrelations in the stream of experienceβ via coding, but the βhowβ is yet to be determined.
(Meanwhile, Outward Bound was a good learning experience for my child. Lightning had them on their backpacks as small as possible. Just glad to know this after the fact.)
Thanks for this Rob. I think you're absolutely right that we need to pay more attention to what companies are involved in the AI push. There are some interesting efforts to try to develop AI models that are made with data from the public commons, and are open source, though I can't say I have high hopes they'll be able to compete with Big Tech.
I tend to go on a bit about this, but small open models trained on curated data stores seem to be a viable future for LLMs as educational tools. The bubbles of AGI hype and AI investment deflating could make that future more conceivable.
You hit the nail on the head here about how the corporate personalities are really starting to show themselves and reveal the desperate profit motivations behind the touted AI βeducationalβ benefits on offer. I do think we will look back on this moment perhaps not with incredulity (mistakes with social media and kids are now obvious in retrospect) but something like criminal negligence that we watched the same thing happen again. The Economistβs leader this week is on how AI is rewiring childhood.
My hopeful moments these days are based on a sense that the executives at the tech giants are as confused and worried as the rest of us. I am sure there are some very rich sociopaths in Silicon Valley, but the majority of the managers and engineers and marketers are human beings who want a better world for themselves and their kids.
I think that's true. Hopefully some are paying attention to what's happening in their children's schools. My optimism for how AI can be helpful in the learning process diminishes by the day, not because I don't think there is enormous potential - the things AI has enabled me to do are extraordinary - but because the scope of the problem seems to be so overwhelming and even worse beneath the surface that I just don't think school leaders truly grasp the magnitude of what's already here, let alone what's on the horizon.
Agreed, with the proviso that problem in nearly all its magnitude predates ChatGPT, and will not be solved merely by "solving for AI."
Of course you are correct. Adding AI to the mix may just be the straw that truly breaks the camels back.
I would like a Picard Mode in my AI models so that they would address me as the captain explicitly as a computer, without all the frippery or any "would you like something else, sir."
I like it. One strategy for disanthropomorphizing chat is to make the role play more specific. I have a friend who uses a robot voice any time he talks with Chat. It helps those around him understand he is addressing a computer and not them. I suspect it also helps him remember that what he is doing is quite different from talking to a human.
There is a well-documented history of Silicon Valley CEOs restricting their kids' technology usage. Despite driving the digital age, leaders like Gates and Jobs limited their childrenβs access, and I suspect current leadership does the same. A Silicon Valley paradox? Unfortunately, not every family has the time or money to establish these boundaries or resist the marketing that suggests children need these devices for educational purposes.
The consequences are significant. Devices are not substitutes for parent involvement, struggling to learn something new and challenging, or even children playing together. The decline of children playing outside erodes the mental and physical resilience that comes from real-world interaction.
The tech industry is now making AI the next frontier, which risks creating a generation unable to learn the fundamentals of their chosen fields. You know by working to understand concepts and solving problems. If AI provides easy answers, many will take the path of least resistance and enter the real world dependent on the software. If the AI isn't available, they will struggle, much like those who cannot perform arithmetic without a calculator. However, the risk here is significantly higher than in previous technological shifts. While a calculator might replace mental arithmetic, AI attempts to replace a broad spectrum of cognitive abilitiesβfrom writing and solving problems to coding and critical analysis.
I am not against AI, but it must be used correctly: apply yourself, struggle to understand what works, and use the tool as a tutor rather than an answer provider. Spending $100k to $400k in college on a college education is a waste of 18 years if you don't know the basics or do not know how to write, analyze, and solve simple problems. If the goal is merely a degree without learning, we should not be surprised when graduates struggle to find employment. The technology is here to stay, so we must teach future generations to use it for productivity, rather than letting it think for them.
Parents also play a significant role in ensuring that children are not focused solely on a piece of paper or specific grades, but instead put themselves through the intellectual struggle that defines real education. This is how I was raised, and it turned out fineβthough I admit it is easier said than done now than it was a few decades ago.
βintellectual struggle that defines real educationβ well said. The deeper the knowledge, the more polymathic cross referencing possiblities arise to meet awareness. Every βhowβ needs a βwhyβ. To choose why we outsource learning to leaning on AI would be to separate flatlined human thinking, quick answers to questions, to more deeply answered dilemmas with steps of probity that include inferences from a lived life of experiences. Holistic thinking is lived learning complexity, not just programmed computing. Perhaps we will get there as in William Jamesβ βrelations in the stream of experienceβ via coding, but the βhowβ is yet to be determined.
(Meanwhile, Outward Bound was a good learning experience for my child. Lightning had them on their backpacks as small as possible. Just glad to know this after the fact.)
Thanks for this Rob. I think you're absolutely right that we need to pay more attention to what companies are involved in the AI push. There are some interesting efforts to try to develop AI models that are made with data from the public commons, and are open source, though I can't say I have high hopes they'll be able to compete with Big Tech.
I tend to go on a bit about this, but small open models trained on curated data stores seem to be a viable future for LLMs as educational tools. The bubbles of AGI hype and AI investment deflating could make that future more conceivable.