Thank you brother. So in my next post I’m going to cover Bina and Daat, but here’s the short answer:
AI is very good at the computation and increasingly good at putting it into words, but what it lacks is the Khokhma, Bina and Daat that exist at the center of that computation in a human mind.
AI can research any question but it lacks the curiosity and interest to ask and pursue a question in the first place that characterizes Khokhma.
AI can deconstruct and reconstruct arguments and extrapolate their implications, which are characteristic of Bina but this process does not deepen its beliefs or beliefs enrich its inner world because it doesn’t have either.
AI is a powerful tool for all of the above, and can help us sift through possibilities of what we should do in our lives, but since it doesn’t have a life itself, it does not and will not have the Daat that you and I have. Any sense on our parts that it does is a hallucination of our own.
There is a kind of wisdom that AI can’t touch- which is the drop of wisdom that God puts into a persons head at just the right time because he wants that flash of insight to be in the world, he wants that person to take a certain action, he wants that soul to complete its mission and transform. We are not brains walking around in a cold world. We are souls with a Mission to do here.
Silicon Valley thinks the world is a gigantic math problem. It’s so much more! And of course that’s why the Torah is multi-layered. We move in thought, speech and action. A thought is one part. Speach and action bring it into the world.
And at the same time, if that action is a mitzvah, there is an additional layer: it’s bringing connection with the divine, drawing something above this world into this world.
Only LaShon HaKodesh can contain the energetic wholeness of any G-dly concept. I touch on this in this talk here - and relate it to what happened at the Tower of Bavel and the Garden of Eden.
Torah is Divine Wisdom so it stands above the strictures of this physical reality. As G-d's wisdom is so far and above our own, whenever ideas are diffused into any other language (eg English), they are only partially whole. G-d recognized the power of language and took this wholeness away from those who wanted to challenge him. Only Lashon HaKodesh can capture the fullness of any idea, and the letters themselves are containers for the spiritual DNA of the concepts they contain (hence, why Adam HaRishon was tasked with naming the animals in the Garden of Eden, essentially meaning, he matched the spiritual DNA demarcated by the letters with the essence of the creatures thereby identifying their true essence).
This may be why it appears that the commentaries contrast each other, when in reality, just as certain Hebrew words can have multiple English definitions, the commentaries which appear to contrast each other, are just getting at different faces of the same truth/idea/reality. The Torah itself is obviously not a "book" in the classical sense. It is more like a painting, a container, into which G-d poured his essence, the into the letters, but also, into the not-letters, and this is also why the Torah operates on multiple levels, black fire on white fire, sound, space, cantillation, language, meaning, visual etc. and helps to explain why the Torah and HaShem are one. This is further elucidated as when the Jews received the Torah at Mount Sinai it suffused all of their senses and why a life of Torah necessitates serving G-d with all of our faculties and senses at once as we become entirely engaged and merged with G-d through doing the Torah and Mitzvoth.
As you mention, all the technology is meant to give us a glimpse into the age of Moschiach, when God will be visible in the world. AI for example shows us what it means to speak and create. Thus we can better appreciate Bereisheit.
There’s friction that actually makes you think—working through a contradiction yourself, reconstructing an argument from scratch, being forced to decide between competing readings. That kind is productive. It deepens understanding.
Then there’s friction that just burns mental energy for no real gain: spending ten minutes hunting down a word, squinting at bad fonts, mechanical decoding that has nothing to do with the logic of what you’re learning. That kind gets in the way of actual comprehension.
Removing the second often frees you up for more of the first. The old system had both mixed together, but treating all friction as formative isn’t an argument—it’s nostalgia. Difficulty and depth aren’t the same thing.
The definition of ḥokhmāh has a similar problem. Calling it receptivity and openness is too vague. Classical sources would call indiscriminate openness *petiyut*—naïve receptivity. Real ḥokhmāh is selective: knowing what deserves attention, what actually matters, what has consequences. Even if binah is supposed to add structure later, you can’t reduce ḥokhmāh to generic openness without creating conceptual problems down the line.
Which is ironic, because that loose definition actually helps AI’s case. AI is perfectly receptive and has zero skin in the game. The real human edge isn’t curiosity—it’s judgment shaped by values and stakes. You hint at this later, but the opening definition undercuts your own point.
And look, you’re right that there’s a real problem here. But then you stop at “we need to train people to value wisdom.” Okay, but how? What does that look like in practice? L’maaseh, I’m not left with anything I can actually do.
Deej, thank you for these very helpful insights. I really appreciate the distinction you’re making between productive friction that leads to personal edification and unproductive friction that just produces useless thermal energy. I’m with you.
To be clear, the purpose of my post was not a Luddite rejection of AI usage, but rather a clarification of the disposition towards wisdom that must be cultivated or else lost completely in these AI times. I work in a high school and I am seeing smart teenagers, taking multiple APs, applying to top schools — who cannot distinguish between a ChatGPT dvar Torah and a good dvar Torah. They don’t use AI to diminish unproductive friction and increase productive friction. They use it as much as they can to get the results they want. This is my concern.
I don’t see the relationship between enabling one not to think in order to produce an output and indifference to quality (dvarim). In fact, one could argue the opposite.
Not being able to distinguish between high quality dvarim and low quality dvarim is either a function of A) low IQ OR B) not enough exposure to the delta or contrast between the two. A ChatGPT dvar can be high quality btw, it’s just not likely without iterations of refinements led by a high IQ prompter. Not sure what GPT usage has to do with it…
David posted a comment that was accidentally deleted. He wrote that one can see how unintelligent AI is by asking it to display a clock face that showed the time as 10 past 11
I think the gaps in this form of intelligence will be closed if they haven’t already. My question is are there forms of intelligence that are uniquely human and cannot be imitated by AI?
Another great substack! Thank you, Rabbi!!
Is it fair to say that AI is skipping ahead to bina and da’at without the ability to possess khokhma?
Thank you brother. So in my next post I’m going to cover Bina and Daat, but here’s the short answer:
AI is very good at the computation and increasingly good at putting it into words, but what it lacks is the Khokhma, Bina and Daat that exist at the center of that computation in a human mind.
AI can research any question but it lacks the curiosity and interest to ask and pursue a question in the first place that characterizes Khokhma.
AI can deconstruct and reconstruct arguments and extrapolate their implications, which are characteristic of Bina but this process does not deepen its beliefs or beliefs enrich its inner world because it doesn’t have either.
AI is a powerful tool for all of the above, and can help us sift through possibilities of what we should do in our lives, but since it doesn’t have a life itself, it does not and will not have the Daat that you and I have. Any sense on our parts that it does is a hallucination of our own.
https://www.instagram.com/reels/DQk95xSCfwC/
Hi Rabbi Jack! I appreciate you and your writing.
I’d like to explore the following ideas with you:
There is a kind of wisdom that AI can’t touch- which is the drop of wisdom that God puts into a persons head at just the right time because he wants that flash of insight to be in the world, he wants that person to take a certain action, he wants that soul to complete its mission and transform. We are not brains walking around in a cold world. We are souls with a Mission to do here.
Silicon Valley thinks the world is a gigantic math problem. It’s so much more! And of course that’s why the Torah is multi-layered. We move in thought, speech and action. A thought is one part. Speach and action bring it into the world.
And at the same time, if that action is a mitzvah, there is an additional layer: it’s bringing connection with the divine, drawing something above this world into this world.
Only LaShon HaKodesh can contain the energetic wholeness of any G-dly concept. I touch on this in this talk here - and relate it to what happened at the Tower of Bavel and the Garden of Eden.
Torah is Divine Wisdom so it stands above the strictures of this physical reality. As G-d's wisdom is so far and above our own, whenever ideas are diffused into any other language (eg English), they are only partially whole. G-d recognized the power of language and took this wholeness away from those who wanted to challenge him. Only Lashon HaKodesh can capture the fullness of any idea, and the letters themselves are containers for the spiritual DNA of the concepts they contain (hence, why Adam HaRishon was tasked with naming the animals in the Garden of Eden, essentially meaning, he matched the spiritual DNA demarcated by the letters with the essence of the creatures thereby identifying their true essence).
This may be why it appears that the commentaries contrast each other, when in reality, just as certain Hebrew words can have multiple English definitions, the commentaries which appear to contrast each other, are just getting at different faces of the same truth/idea/reality. The Torah itself is obviously not a "book" in the classical sense. It is more like a painting, a container, into which G-d poured his essence, the into the letters, but also, into the not-letters, and this is also why the Torah operates on multiple levels, black fire on white fire, sound, space, cantillation, language, meaning, visual etc. and helps to explain why the Torah and HaShem are one. This is further elucidated as when the Jews received the Torah at Mount Sinai it suffused all of their senses and why a life of Torah necessitates serving G-d with all of our faculties and senses at once as we become entirely engaged and merged with G-d through doing the Torah and Mitzvoth.
As you mention, all the technology is meant to give us a glimpse into the age of Moschiach, when God will be visible in the world. AI for example shows us what it means to speak and create. Thus we can better appreciate Bereisheit.
Have you seen this shiur by YY Kazen, founder of Chabad.org: https://www.chabad.org/library/article_cdo/aid/2436505/jewish/Introducing-Chabad-Lubavitch-in-Cyberspace.htm
You’re mixing up two different kinds of friction.
There’s friction that actually makes you think—working through a contradiction yourself, reconstructing an argument from scratch, being forced to decide between competing readings. That kind is productive. It deepens understanding.
Then there’s friction that just burns mental energy for no real gain: spending ten minutes hunting down a word, squinting at bad fonts, mechanical decoding that has nothing to do with the logic of what you’re learning. That kind gets in the way of actual comprehension.
Removing the second often frees you up for more of the first. The old system had both mixed together, but treating all friction as formative isn’t an argument—it’s nostalgia. Difficulty and depth aren’t the same thing.
The definition of ḥokhmāh has a similar problem. Calling it receptivity and openness is too vague. Classical sources would call indiscriminate openness *petiyut*—naïve receptivity. Real ḥokhmāh is selective: knowing what deserves attention, what actually matters, what has consequences. Even if binah is supposed to add structure later, you can’t reduce ḥokhmāh to generic openness without creating conceptual problems down the line.
Which is ironic, because that loose definition actually helps AI’s case. AI is perfectly receptive and has zero skin in the game. The real human edge isn’t curiosity—it’s judgment shaped by values and stakes. You hint at this later, but the opening definition undercuts your own point.
And look, you’re right that there’s a real problem here. But then you stop at “we need to train people to value wisdom.” Okay, but how? What does that look like in practice? L’maaseh, I’m not left with anything I can actually do.
Deej, thank you for these very helpful insights. I really appreciate the distinction you’re making between productive friction that leads to personal edification and unproductive friction that just produces useless thermal energy. I’m with you.
To be clear, the purpose of my post was not a Luddite rejection of AI usage, but rather a clarification of the disposition towards wisdom that must be cultivated or else lost completely in these AI times. I work in a high school and I am seeing smart teenagers, taking multiple APs, applying to top schools — who cannot distinguish between a ChatGPT dvar Torah and a good dvar Torah. They don’t use AI to diminish unproductive friction and increase productive friction. They use it as much as they can to get the results they want. This is my concern.
I don’t see the relationship between enabling one not to think in order to produce an output and indifference to quality (dvarim). In fact, one could argue the opposite.
Not being able to distinguish between high quality dvarim and low quality dvarim is either a function of A) low IQ OR B) not enough exposure to the delta or contrast between the two. A ChatGPT dvar can be high quality btw, it’s just not likely without iterations of refinements led by a high IQ prompter. Not sure what GPT usage has to do with it…
I think you’re missing one more possibility:
C) Not caring to learn what the distinction is.
This is what I’m suggesting should be especially cultivated in times when the option to not think at all has been so greatly facilitated.
David posted a comment that was accidentally deleted. He wrote that one can see how unintelligent AI is by asking it to display a clock face that showed the time as 10 past 11
I think the gaps in this form of intelligence will be closed if they haven’t already. My question is are there forms of intelligence that are uniquely human and cannot be imitated by AI?