While explaining *Ashtavakra Gita* to his students, Osho said "This is one of the best and beautiful scriptures but don't get trapped in its words either. They are beautiful, not because they hold the truth but because they were spoken by someone who had Truth within him."
He further added, "Truth is in silence. Once spoken, it is no longer the truth. No shabda (word) can hold the truth. Still many scriptures were written in an attempt to write the Truth, knowing very well it will fail. Here, I speak to you not to tell you the truth but to create the hunger within you that will seek the Truth. My words act like *ghee* that will ignite the fire within you."
Before I follow the bread crumbs at the bottom, I want to express my joy at the phrase "epistemic struggle"; and "facile knowledge" caught my eye as I scrolled back, too. They feel like very good labels for ideas I've groped and stumbled around a lot lately
My long-held criticism of things (and the people who readily cling to such things) such as Roman Catholic catechism books and ChatGPT is that one is not left to formulate, and then to wrestle with the question(s); which is, for me, at least, the most precious part of the process.
Certainty is over-rated and over-valued almost beyond calculation in our efficiency-worshipping culture.
I feel like some books are sacred in a way. A person's knowledge and life experiences crystallised with their own words. It's like you skip all the usual levels of conversation and go straight to the depth of their minds. I've cried over books and art often, I think it's natural to be so moved by such pure emotion when we walk around with our polite and practiced selves in proper society.
The other issue with summaries and additional layers of abstraction is that they frequently leave out important details and context assuming that they aren’t factually wrong. There is a reason why children were once told to only use primary sources when writing term papers. Even an encyclopedia was too far removed to be considered an unambiguous source.
When you outsource your information retrieval and summarization, you are relinquishing the ability to come to your own conclusions and subjugating your understanding to whatever framework was used to generate the summary.
To expand on this point, this issue is particularly salient with regard to LLMs and their statistical predictions, because they are measuring “truthiness” rather than any kind of objective fact, which explains why they have a tendency to hallucinate. Having said that, the issue is not limited to AI, as there is plenty of “scholarship” whose sole purpose is to defend the indefensible, and if you do not have enough context to evaluate your sources, you will be less knowledgeable for reading it.
A particularly pernicious example of this are the various writers in the post US civil war period whose tracts are fundamentally a justification of treason in defense of slavery and criticism of reconstruction. Some of these false narratives were taught well into the 1980’s. Unfortunately, it’s not always the victors that write the histories.
AI is merely regurgitating a statistical model of the content it has consumed. The quality or veracity of the content isn’t factored.
I appreciate the opinion and it is well written, but to me it feels a little bit as if it is romanticising avoidable frustration.
I also have a PhD and I now work in science, and, in my experience, not being able to find the resources to do research does not primarily lead to catharsis and illumination, but a lot of page-flipping and time-wasting. People are (rightfully) concerned about genAI's impact on knowledge and information, yet I can't recall anyone making the same argument against Google Scholar although they can serve similar purposes: finding and presenting relevant resources to further progress. Without tools such as these, it would take (and did take) a lot of time just to look up things, to find them.
Personally, I think that one of genAI's best use cases is as a search engine. The ability to pose questions in natural language rather than have to guess a combination of keywords is a powerful one.
I do appreciate the perspective, and I do think we need to be vigilant about offloading too much of our mental capacity to machines. However, I also don't think we should idealise and aspire to return to a time where resources are by design and by lack of technology hard to come by and time-consuming to consult.
This is good balance. 'Manual' or 'granular' research practices have their purpose are not 'inherently' superior to high-level familiarity. Sometimes you need to see the forest, sometimes you need to study the tree. I've never read Yoshida, do you think he'd agree?
That was amazing. I was earlier today listening to a podcast by Stephen West and over there he was talking about something similar, how we are able to answer questions today and build things because of the prior knowledge. Sadly, we also have taken that knowledge and the effort to gather it all for granted. And like people say, lost in translation, what are we losing by not putting in the requisite effort?
beware of Greeks bringing gifts (such as Artificial Intelligence) and of Japanese bringing koans to westerners. One can travel in circles for decades and advance not one inch. It's one of the deadly diseases that afflict academia and that inquiring minds can come down with even at a distance of thousands of miles. Best regards to you.--Jim
Very insightful. This is a problem that is only going to intensify with the mass usage of chatbots. Anecdotal, but the quality of my questions have definitely deteriorated the more I use ai for assignments.
AI also has knowledge gaps too, and unlike an expert in their field it's more inclined to make up a feasible answer than admit ignorance. Just like the oracle at Delphi, AI provides the illusion of omniscience. Beyond that there can be great value in not knowing. Selective ignorance is a valuable practice in this information age. There's a profound peace that comes with recognizing we can't know everything and some subjects aren't worth the hours they require to master.
I really look forward to your observations, and this rings true- to know something is not necessarily the facts but the wisdom, and Ai can only draw on what’s available from certain sources, no human intuition there. And totally agree the lack of effort put in to acquiring the knowledge this way means very little sticks in anyone’s long term memory- even not having to learn phone numbers anymore shows us that!
Definitely a part of this ‘needing an immediate answer’ way of thinking comes from our education. Standardized testing requires this type of thinking. You just need to know the answer and spew it back out.
On top of that, I would argue many people are in a ‘get rich quick’ mindset in the current state of things, and that same mantra is applied to knowledge and/or information. We need to know what’s happening as soon as possible to ‘stay ahead of the curve’, which can lead to shortcuts in thinking and info. Media is the most guilty of this.
While explaining *Ashtavakra Gita* to his students, Osho said "This is one of the best and beautiful scriptures but don't get trapped in its words either. They are beautiful, not because they hold the truth but because they were spoken by someone who had Truth within him."
He further added, "Truth is in silence. Once spoken, it is no longer the truth. No shabda (word) can hold the truth. Still many scriptures were written in an attempt to write the Truth, knowing very well it will fail. Here, I speak to you not to tell you the truth but to create the hunger within you that will seek the Truth. My words act like *ghee* that will ignite the fire within you."
Osho, AKA Bhagwan Shree Rajneesh, scammed thousands of followers out of their life savings.
I am sure he kited those words from someone else, as well.
Truth is one, but with many faces. Subjective truth is one's experience and that's it.
Churchill's bust is kept in the White House while he is one of the most hated in India for atrocities he committed here.
Before I follow the bread crumbs at the bottom, I want to express my joy at the phrase "epistemic struggle"; and "facile knowledge" caught my eye as I scrolled back, too. They feel like very good labels for ideas I've groped and stumbled around a lot lately
My long-held criticism of things (and the people who readily cling to such things) such as Roman Catholic catechism books and ChatGPT is that one is not left to formulate, and then to wrestle with the question(s); which is, for me, at least, the most precious part of the process.
Certainty is over-rated and over-valued almost beyond calculation in our efficiency-worshipping culture.
So, Bravo!
Loved it! Thanks to share it :)
I feel like some books are sacred in a way. A person's knowledge and life experiences crystallised with their own words. It's like you skip all the usual levels of conversation and go straight to the depth of their minds. I've cried over books and art often, I think it's natural to be so moved by such pure emotion when we walk around with our polite and practiced selves in proper society.
Wow, that is a wonderful way of describing books. And its absolutely true.
How beautiful! Thanks for introducing us to Professor Yoshida
The other issue with summaries and additional layers of abstraction is that they frequently leave out important details and context assuming that they aren’t factually wrong. There is a reason why children were once told to only use primary sources when writing term papers. Even an encyclopedia was too far removed to be considered an unambiguous source.
When you outsource your information retrieval and summarization, you are relinquishing the ability to come to your own conclusions and subjugating your understanding to whatever framework was used to generate the summary.
To expand on this point, this issue is particularly salient with regard to LLMs and their statistical predictions, because they are measuring “truthiness” rather than any kind of objective fact, which explains why they have a tendency to hallucinate. Having said that, the issue is not limited to AI, as there is plenty of “scholarship” whose sole purpose is to defend the indefensible, and if you do not have enough context to evaluate your sources, you will be less knowledgeable for reading it.
A particularly pernicious example of this are the various writers in the post US civil war period whose tracts are fundamentally a justification of treason in defense of slavery and criticism of reconstruction. Some of these false narratives were taught well into the 1980’s. Unfortunately, it’s not always the victors that write the histories.
AI is merely regurgitating a statistical model of the content it has consumed. The quality or veracity of the content isn’t factored.
I appreciate the opinion and it is well written, but to me it feels a little bit as if it is romanticising avoidable frustration.
I also have a PhD and I now work in science, and, in my experience, not being able to find the resources to do research does not primarily lead to catharsis and illumination, but a lot of page-flipping and time-wasting. People are (rightfully) concerned about genAI's impact on knowledge and information, yet I can't recall anyone making the same argument against Google Scholar although they can serve similar purposes: finding and presenting relevant resources to further progress. Without tools such as these, it would take (and did take) a lot of time just to look up things, to find them.
Personally, I think that one of genAI's best use cases is as a search engine. The ability to pose questions in natural language rather than have to guess a combination of keywords is a powerful one.
I do appreciate the perspective, and I do think we need to be vigilant about offloading too much of our mental capacity to machines. However, I also don't think we should idealise and aspire to return to a time where resources are by design and by lack of technology hard to come by and time-consuming to consult.
This is good balance. 'Manual' or 'granular' research practices have their purpose are not 'inherently' superior to high-level familiarity. Sometimes you need to see the forest, sometimes you need to study the tree. I've never read Yoshida, do you think he'd agree?
That was amazing. I was earlier today listening to a podcast by Stephen West and over there he was talking about something similar, how we are able to answer questions today and build things because of the prior knowledge. Sadly, we also have taken that knowledge and the effort to gather it all for granted. And like people say, lost in translation, what are we losing by not putting in the requisite effort?
I've never thought about that before, how fascinating. Loved Algospeak as well.
Hi Adam,
beware of Greeks bringing gifts (such as Artificial Intelligence) and of Japanese bringing koans to westerners. One can travel in circles for decades and advance not one inch. It's one of the deadly diseases that afflict academia and that inquiring minds can come down with even at a distance of thousands of miles. Best regards to you.--Jim
Very insightful. This is a problem that is only going to intensify with the mass usage of chatbots. Anecdotal, but the quality of my questions have definitely deteriorated the more I use ai for assignments.
AI also has knowledge gaps too, and unlike an expert in their field it's more inclined to make up a feasible answer than admit ignorance. Just like the oracle at Delphi, AI provides the illusion of omniscience. Beyond that there can be great value in not knowing. Selective ignorance is a valuable practice in this information age. There's a profound peace that comes with recognizing we can't know everything and some subjects aren't worth the hours they require to master.
I really look forward to your observations, and this rings true- to know something is not necessarily the facts but the wisdom, and Ai can only draw on what’s available from certain sources, no human intuition there. And totally agree the lack of effort put in to acquiring the knowledge this way means very little sticks in anyone’s long term memory- even not having to learn phone numbers anymore shows us that!
This was such excellent thought and writing. Thank you for it. 🙏
"Figuring out which question to ask is more important than the answer itself."
I'm in.
Definitely a part of this ‘needing an immediate answer’ way of thinking comes from our education. Standardized testing requires this type of thinking. You just need to know the answer and spew it back out.
On top of that, I would argue many people are in a ‘get rich quick’ mindset in the current state of things, and that same mantra is applied to knowledge and/or information. We need to know what’s happening as soon as possible to ‘stay ahead of the curve’, which can lead to shortcuts in thinking and info. Media is the most guilty of this.