Thanks for 2 things, the bit on AI Bing and the link to Kareem's piece. On the first, as a school librarian, I teach research skills to grades 6-9. How will this affect the research paper assignment? English and History teachers are already worried about how Chat GPT will effect the writing of the research paper. Now librarians will have to worry about how AI will effect the research part of the paper.
On Kareem, I sent his piece to a long-time friend who turned me on to Kareem's sky hook when he was playing for the Lakers in the 80s. He'll appreciate it, I'm sure.
I'm a software engineer but I don't work in machine learning. I'm probably better versed on the subject than most but far from a subject matter expert. If anything I say gets corrected by an expert I'll certainly defer to that person.
With that, what we're seeing is machine learning and that's something that can be really useful. ChatGPT is more of a gimmick or novelty. When I logged into the company's website as a developer they had quite a few other things to show and I'm sure many would be useful.
Machine learning can also be labeled Narrow AI. Kind of like it sounds, it's set up for one particular task like being a chatbot or reading radiology images. We've all heard of the biases that find their way in because of the datasets used to train them.
What people fear is General AI. That's a complex issue and probably at least decades away because it would me something is sentient. The first question is why would we ever develop General AI? We have billions of them on the planet already and truthfully it makes more sense to give human brains additional tools to enhance what they can do instead of trying to build another sentient being. There's the risk and then the ethical questions. We could save ourselves a lot of trouble and stick with a collection of narrow AI to help us out.
I can envision yet another arms race between Narrow AI systems designed to sell us stuff and other Narrow AI that helps us filter it out. We tend to become pretty good at that anyway. We can all read web sites and tune out the ads or ignore them in any feeds we have. It wasn't that long ago that people feared there were subliminal messages in cinema ads.
Some people won't be able to distinguish between genuine content and fake content - like what happens now. This is the Next Level (so to speak).
I think other things are a bigger threat now, like the Republican Party as it stands or an errant nuclear exchange. A search engine finding a new way to stear people isn't high on my list of concerns.
I could be.wrong. I have built enough software in my career to know how badly so much software is written. One thing that should worry is all is how many critical systems are still in COBOL and how many COBOL programmers keep retiring in large numbers.
My position has always been that (i) AI should be heavily regulated, (ii) is a National Security issue, and (iii) therefore, the government, and not the private sector, should be the head honcho in this. As such, similar to our other defense related industries, the Federal Government should be the only allowed buyer of AI technologies; we should not allow AI to be used by private companies in any way, shape or form. Doing this would both (i) insure that investments in AI continue, as a way to protect against China and Iranian deployments of AI, (ii) while at the same time, insuring that AI is not used by private companies to not only replace human beings in the workforce, which is the plan, but to also put their competitors out of business.
Anyone who thinks that there is some positive benefit to the development of fully functional AI that does not result in mass unemployment/dislocation, and even more concentration of wealth is living in a fantasy world. I would advise that those people read the Dune series of books (Butlerian Jihad), especially the prequels regarding the "enslavement" of man by machines. If you understand the book, it wasn't that machines actually enslaved people a la Battlestar Galactica, it was that the owners of the machines did. Based on what we are seeing with these tech companies now, can anyone say with a straight face that this is not happening already?
Personally, I am not a fan of ChatGPT and most AI in general. I look at it and see the Law of Unintended Consequences. I personally am in favor of HEAVY regulation of AI. I see some very negative things coming. I am not being Debbie Downer about this, but bad things will come from the growing acceptance of AI.
"If you have had any contact with the MAGA section of the Catholic Church over the last seven years, then you have some sense of how absolutely forking insane Pavone had to be to get laicized. Because the Church has basically been a Trump Party free-for-all not all that different from the Southern Baptist Convention."
I'm not Catholic but I did attend law school at a Jesuit university. The closest we ever got to any religion though was a brief prayer at the graduation ceremony that began "God of inclusion." The university as a whole seems to have made its peace with having LBGTQ students and organizations as well. Maybe this is a location thing but I always thought that Catholics as a whole were pretty split but a lot of the American leadership is more reactionary.
I’m an atheist liberal and I don’t always agree with what is said on the Bulwark, but I would never cancel my subscription over something that offended me. This is the difference between the intellectually curious and those other people. You sit somewhere between my view of the world and Republican’s view,if it can be called that. I find it interesting,informative,and sometimes frustrating but always entertaining. I’m 77 and have lived in europe the past 15 years and never subscribed to a podcast,but yours perked my interest. I’m glad to see other subjects discussed besides Trump! Best regards and I love your facial contortions.😊
'Because the Church has basically been a Trump Party free-for-all not all that different from the Southern Baptist Convention.'
I wish you had specified the 'US Church' has been a free for all not that different from the Southern Convention. Two reasons for that JV, that US connection has been going on for 25+ years and the second is, Pope Francis has not endorsed that connection and has actually taught a kinder gentler approach.
I've been watching "Air Disasters" on Smithsonian Channel - as if I needed another reason not to fly! What surprises me is how many accidents are a result of computer/technical errors - mechanical glitches or plain old computer glitches - that the humans either didn't anticipate or can't resolve because of the complexity of the machines. In other words, screw self-driving cars/planes, etc. They actually need the humans, faulty as we are, to operate them.
Except that there's always room for human error. That computer "glitch" might be made by a programmer who hadn't had his morning coffee. But very often, it is a cascade of human error that brings a plane down. For your reading pleasure, many such stories can be found here:
One of the biggest mistakes the US made was dumping a great passenger RR system for cars. I love watching "Mighty Trains" on Smithsonian Channel. Even today, trains provide substantial percentages of moving goods in the US.
The track record for AI is not great. Predictive analytics is unfair to many people. For instance, the predictive algorithms that determine decisions about prison recidivism and leniency tend to be twice as hard on black men as on white men. (One of many examples, and not all of the bad ones are in the US or China. There's a horrific case of misplaced fraud crackdowns in the Netherlands. Biased decisions, using AI as justification, took children away from caring parents.)
One of the many ironies of predictive analytics is that it's hard to know, in advance, whether it will be accurate.
The problem with AI is indeed GIGO, but with a twist. It gives people a feeling that we all crave, that it's human nature to crave: the feeling of being in control of things. Meanwhile, it makes it harder to know when you have genuine control and when you don't.
People who are supposed to oversee automated systems tend to slough off. This phenomenon was first noticed in airplane pilot training. It's called "automated complacency."
Delegating tasks to AI, like having it compose your email, gives you a feeling of control, a little rush. You're getting it done! But you've actually given up a little piece of control. This is called "frictionlessness."
Automated complacency and frictionlessness are addictive. Unfortunately, the only solution is to pay attention to when you delegate, and keep monitoring it. Which takes away from that terrific feeling.
Or write fewer emails. And already, you're probably not keeping up. I'm not.
As Sheena Iyengar says, You know you have real control over something when it's difficult.
Juliette Powell and I have a book coming out this summer called The AI Dilemma that goes into all this. ChatGPT and the new Bing appeared just in time for us to include them in the book. But the real dilemma, the question of how to live well with AI, is not about the latest release. It's not even about AI technology.
The tech is a forcing function. It's forcing us to develop the skill and fortitude we would need anyway -- among other things, to deal with politics. Not sure we're going to develop it in time, but I tend to buy Thornton Wilder's optimistic view of human nature in The Skin of Our Teeth.
'The tech is a forcing function. It's forcing us to develop the skill and fortitude we would need anyway -- among other things, to deal with politics."
Love this, Trump was also a forcing function. I hope we and the Bulwark folks have developed the skill and fortitude we need to deal with Trump and his followers....including Desantis.
Three cheers for The Pillar (with an extra cheer for Mrs Last's Twitter timeline since they published the first Pavone article . . . though I don't envy her mentions right about now) . . . (also I'm *still* trying to quit Twitter) . . .
What a wonderful article about Kareem Abdul Jabbar. As you 24 year old, I got to see Lew Alcindor play in the UCLA basketball stadium. Our seats were so high up, I could almost touch the ceiling. He was a genius at the game of basketball and, it would seem, a genius at the game of life.
I probably would be a Catholic today if my grandmother hadn’t left the church because of an attempted sexual assault by a priest. I didn't know this until after she had passed away, but some of her faith was always in me, and I have been infatuated with Catholicism as long as I can remember. As a child, I begged my parents to let me go church with my Catholic friends; I was completely enrapt when I discovered the writings of St. Teresa of Avila; my proudest moment was a getting graduate degree from a Jesuit intuition. And yet….the things that have always stopped me in the thousand times I contemplated becoming an actual member of the Church were the ever-present reminders of the near zero tolerance for abortion and the seemingly unlimited tolerance for sexual abuse. Makes me sad when I see those two things entangled into one big mess after another.
Hahaha! No papers required. There is a wonderful book called “Searching for Sunday” by Rachel Held Evans that described what it was like when she started going to an Episcopal church. People are low key friendly but will probably leave you be—pushiness isn’t our style. We offer communion to all baptized Christians, so we do get a lot of Catholics who for whatever reason aren’t receiving there.
"(Karem) Tied with Mary Carillo as the most intellectually interesting professional athletes of the last half-century" So by that you mean after 1973?
Bill Russell was coaching the Seattle Supersonics till '77 and coached in the '80s. He's my vote for the most intellectually interesting professional athlete in American history.
Other than being a big-time basketball player "before my time", I knew nothing about Russell when I saw him speak around 1980 when I was in college. I wasn't expecting a jock would teach me a life changing worldview that night.
He basically gave a talk about the difference between people acting in their immediate self interest versus acting on their long-term "enlighten self interest". Only by acting with enlighten self interest can we ourselves be "free". By tolerating injustice of others--acting out of short-term self-interest--we will never be free ourselves. His example was of course centered around racism and why it was in the enlighten self-interest of the predominately white middle-class audience he was speaking to, to support the civil rights movement.
I'd like to connect that lesson I learned from Russell to (some of ) your writings and especially the GREAT podcast "two wolves" you and Sarah had on Friday. It is in our short term interest to treat the wolf as an enemy that we should fear (and the wolf will act accordingly). But it is in our enlighten self interest to feed the wolf and give it respect.
In terms of your writing, you preach to "love" others, ie Cletus the Trumper. While I have little sympathy for Trumpers--and I've tried--I recognize you are correct (and I give you a pass on your occasional mocking of Cletus) that we need to understand what makes Cletus tick. It is in my enlightened self-interest to try and understand Cletus.
Finally, the best Bill Russell story is why he often refused to sign autographs, as he'd rather shake your hand and have a short conversation.
Bill Russell wrote an autobiography, and for decades wrote a column, not a sports column but whatever he was interested in. I think you would like him.
Thanks for 2 things, the bit on AI Bing and the link to Kareem's piece. On the first, as a school librarian, I teach research skills to grades 6-9. How will this affect the research paper assignment? English and History teachers are already worried about how Chat GPT will effect the writing of the research paper. Now librarians will have to worry about how AI will effect the research part of the paper.
On Kareem, I sent his piece to a long-time friend who turned me on to Kareem's sky hook when he was playing for the Lakers in the 80s. He'll appreciate it, I'm sure.
Pavone just looks like a pervert. Thanks for recommending The Pillar. Went there. Looks good. Will subscribe.
I'm a software engineer but I don't work in machine learning. I'm probably better versed on the subject than most but far from a subject matter expert. If anything I say gets corrected by an expert I'll certainly defer to that person.
With that, what we're seeing is machine learning and that's something that can be really useful. ChatGPT is more of a gimmick or novelty. When I logged into the company's website as a developer they had quite a few other things to show and I'm sure many would be useful.
Machine learning can also be labeled Narrow AI. Kind of like it sounds, it's set up for one particular task like being a chatbot or reading radiology images. We've all heard of the biases that find their way in because of the datasets used to train them.
What people fear is General AI. That's a complex issue and probably at least decades away because it would me something is sentient. The first question is why would we ever develop General AI? We have billions of them on the planet already and truthfully it makes more sense to give human brains additional tools to enhance what they can do instead of trying to build another sentient being. There's the risk and then the ethical questions. We could save ourselves a lot of trouble and stick with a collection of narrow AI to help us out.
I can envision yet another arms race between Narrow AI systems designed to sell us stuff and other Narrow AI that helps us filter it out. We tend to become pretty good at that anyway. We can all read web sites and tune out the ads or ignore them in any feeds we have. It wasn't that long ago that people feared there were subliminal messages in cinema ads.
Some people won't be able to distinguish between genuine content and fake content - like what happens now. This is the Next Level (so to speak).
I think other things are a bigger threat now, like the Republican Party as it stands or an errant nuclear exchange. A search engine finding a new way to stear people isn't high on my list of concerns.
I could be.wrong. I have built enough software in my career to know how badly so much software is written. One thing that should worry is all is how many critical systems are still in COBOL and how many COBOL programmers keep retiring in large numbers.
My position has always been that (i) AI should be heavily regulated, (ii) is a National Security issue, and (iii) therefore, the government, and not the private sector, should be the head honcho in this. As such, similar to our other defense related industries, the Federal Government should be the only allowed buyer of AI technologies; we should not allow AI to be used by private companies in any way, shape or form. Doing this would both (i) insure that investments in AI continue, as a way to protect against China and Iranian deployments of AI, (ii) while at the same time, insuring that AI is not used by private companies to not only replace human beings in the workforce, which is the plan, but to also put their competitors out of business.
Anyone who thinks that there is some positive benefit to the development of fully functional AI that does not result in mass unemployment/dislocation, and even more concentration of wealth is living in a fantasy world. I would advise that those people read the Dune series of books (Butlerian Jihad), especially the prequels regarding the "enslavement" of man by machines. If you understand the book, it wasn't that machines actually enslaved people a la Battlestar Galactica, it was that the owners of the machines did. Based on what we are seeing with these tech companies now, can anyone say with a straight face that this is not happening already?
Personally, I am not a fan of ChatGPT and most AI in general. I look at it and see the Law of Unintended Consequences. I personally am in favor of HEAVY regulation of AI. I see some very negative things coming. I am not being Debbie Downer about this, but bad things will come from the growing acceptance of AI.
Not related to this post but just to say that I do think your negative energy is why the Eagles lost. I hate it when JVL is right.
"If you have had any contact with the MAGA section of the Catholic Church over the last seven years, then you have some sense of how absolutely forking insane Pavone had to be to get laicized. Because the Church has basically been a Trump Party free-for-all not all that different from the Southern Baptist Convention."
I'm not Catholic but I did attend law school at a Jesuit university. The closest we ever got to any religion though was a brief prayer at the graduation ceremony that began "God of inclusion." The university as a whole seems to have made its peace with having LBGTQ students and organizations as well. Maybe this is a location thing but I always thought that Catholics as a whole were pretty split but a lot of the American leadership is more reactionary.
I’m an atheist liberal and I don’t always agree with what is said on the Bulwark, but I would never cancel my subscription over something that offended me. This is the difference between the intellectually curious and those other people. You sit somewhere between my view of the world and Republican’s view,if it can be called that. I find it interesting,informative,and sometimes frustrating but always entertaining. I’m 77 and have lived in europe the past 15 years and never subscribed to a podcast,but yours perked my interest. I’m glad to see other subjects discussed besides Trump! Best regards and I love your facial contortions.😊
🤪
'Because the Church has basically been a Trump Party free-for-all not all that different from the Southern Baptist Convention.'
I wish you had specified the 'US Church' has been a free for all not that different from the Southern Convention. Two reasons for that JV, that US connection has been going on for 25+ years and the second is, Pope Francis has not endorsed that connection and has actually taught a kinder gentler approach.
I should have said that explicitly. I think I’m an earlier draft I did but then cut it in my constant attempts to save words. Sorry.
I've been watching "Air Disasters" on Smithsonian Channel - as if I needed another reason not to fly! What surprises me is how many accidents are a result of computer/technical errors - mechanical glitches or plain old computer glitches - that the humans either didn't anticipate or can't resolve because of the complexity of the machines. In other words, screw self-driving cars/planes, etc. They actually need the humans, faulty as we are, to operate them.
Except that there's always room for human error. That computer "glitch" might be made by a programmer who hadn't had his morning coffee. But very often, it is a cascade of human error that brings a plane down. For your reading pleasure, many such stories can be found here:
https://admiralcloudberg.medium.com/
I'll take the train. :)
One of the biggest mistakes the US made was dumping a great passenger RR system for cars. I love watching "Mighty Trains" on Smithsonian Channel. Even today, trains provide substantial percentages of moving goods in the US.
The track record for AI is not great. Predictive analytics is unfair to many people. For instance, the predictive algorithms that determine decisions about prison recidivism and leniency tend to be twice as hard on black men as on white men. (One of many examples, and not all of the bad ones are in the US or China. There's a horrific case of misplaced fraud crackdowns in the Netherlands. Biased decisions, using AI as justification, took children away from caring parents.)
One of the many ironies of predictive analytics is that it's hard to know, in advance, whether it will be accurate.
The problem with AI is indeed GIGO, but with a twist. It gives people a feeling that we all crave, that it's human nature to crave: the feeling of being in control of things. Meanwhile, it makes it harder to know when you have genuine control and when you don't.
People who are supposed to oversee automated systems tend to slough off. This phenomenon was first noticed in airplane pilot training. It's called "automated complacency."
Delegating tasks to AI, like having it compose your email, gives you a feeling of control, a little rush. You're getting it done! But you've actually given up a little piece of control. This is called "frictionlessness."
Automated complacency and frictionlessness are addictive. Unfortunately, the only solution is to pay attention to when you delegate, and keep monitoring it. Which takes away from that terrific feeling.
Or write fewer emails. And already, you're probably not keeping up. I'm not.
As Sheena Iyengar says, You know you have real control over something when it's difficult.
Juliette Powell and I have a book coming out this summer called The AI Dilemma that goes into all this. ChatGPT and the new Bing appeared just in time for us to include them in the book. But the real dilemma, the question of how to live well with AI, is not about the latest release. It's not even about AI technology.
The tech is a forcing function. It's forcing us to develop the skill and fortitude we would need anyway -- among other things, to deal with politics. Not sure we're going to develop it in time, but I tend to buy Thornton Wilder's optimistic view of human nature in The Skin of Our Teeth.
'The tech is a forcing function. It's forcing us to develop the skill and fortitude we would need anyway -- among other things, to deal with politics."
Love this, Trump was also a forcing function. I hope we and the Bulwark folks have developed the skill and fortitude we need to deal with Trump and his followers....including Desantis.
Three cheers for The Pillar (with an extra cheer for Mrs Last's Twitter timeline since they published the first Pavone article . . . though I don't envy her mentions right about now) . . . (also I'm *still* trying to quit Twitter) . . .
What a wonderful article about Kareem Abdul Jabbar. As you 24 year old, I got to see Lew Alcindor play in the UCLA basketball stadium. Our seats were so high up, I could almost touch the ceiling. He was a genius at the game of basketball and, it would seem, a genius at the game of life.
I probably would be a Catholic today if my grandmother hadn’t left the church because of an attempted sexual assault by a priest. I didn't know this until after she had passed away, but some of her faith was always in me, and I have been infatuated with Catholicism as long as I can remember. As a child, I begged my parents to let me go church with my Catholic friends; I was completely enrapt when I discovered the writings of St. Teresa of Avila; my proudest moment was a getting graduate degree from a Jesuit intuition. And yet….the things that have always stopped me in the thousand times I contemplated becoming an actual member of the Church were the ever-present reminders of the near zero tolerance for abortion and the seemingly unlimited tolerance for sexual abuse. Makes me sad when I see those two things entangled into one big mess after another.
The Episcopal church welcomes you. ;-)
Is it that simple? Can I just show up? I don't need papers, or months of classes or something?
After getting married and having my son baptized Catholic, I feel like I should bring some papers.
Hahaha! No papers required. There is a wonderful book called “Searching for Sunday” by Rachel Held Evans that described what it was like when she started going to an Episcopal church. People are low key friendly but will probably leave you be—pushiness isn’t our style. We offer communion to all baptized Christians, so we do get a lot of Catholics who for whatever reason aren’t receiving there.
"Searching for Sunday" is one of my favorite books!
JVL
I believe you’re an exponential level of concern too low. It’s not a 17, it’s 17 squared.
I think 2001 A Space Odyssey is appropriate; “I'm sorry Dave, I'm afraid I can't do that."
"(Karem) Tied with Mary Carillo as the most intellectually interesting professional athletes of the last half-century" So by that you mean after 1973?
Bill Russell was coaching the Seattle Supersonics till '77 and coached in the '80s. He's my vote for the most intellectually interesting professional athlete in American history.
Maybe? I don’t know enough about him. But Mary and Kareem are writers and I love that.
Other than being a big-time basketball player "before my time", I knew nothing about Russell when I saw him speak around 1980 when I was in college. I wasn't expecting a jock would teach me a life changing worldview that night.
He basically gave a talk about the difference between people acting in their immediate self interest versus acting on their long-term "enlighten self interest". Only by acting with enlighten self interest can we ourselves be "free". By tolerating injustice of others--acting out of short-term self-interest--we will never be free ourselves. His example was of course centered around racism and why it was in the enlighten self-interest of the predominately white middle-class audience he was speaking to, to support the civil rights movement.
I'd like to connect that lesson I learned from Russell to (some of ) your writings and especially the GREAT podcast "two wolves" you and Sarah had on Friday. It is in our short term interest to treat the wolf as an enemy that we should fear (and the wolf will act accordingly). But it is in our enlighten self interest to feed the wolf and give it respect.
In terms of your writing, you preach to "love" others, ie Cletus the Trumper. While I have little sympathy for Trumpers--and I've tried--I recognize you are correct (and I give you a pass on your occasional mocking of Cletus) that we need to understand what makes Cletus tick. It is in my enlightened self-interest to try and understand Cletus.
Finally, the best Bill Russell story is why he often refused to sign autographs, as he'd rather shake your hand and have a short conversation.
Bill Russell wrote an autobiography, and for decades wrote a column, not a sports column but whatever he was interested in. I think you would like him.