"It is fun to play with and there are many possible applications." (referring to ChatGPT)
Not sure if I disagree or simply misunderstand your comment yesterday about delving into WWW and ChatGPT. I agree with this one.,
As for Jaron’s claim about fakes, I'm skeptical it matters other than tangential questions such as pinching entry level jobs, flooding requests for comment, and use akin to plagarism in school. Or are those your concerns? My concerns regarding writing are whether it is clearly wrong factually and if it tells a compelling story. I don't think it's necessary to fine tune or expand on that to say my concerns are no different than for entirely human composition. "Fake" seems to be a misguided put down akin to luddite technophobia. But maybe that's too harsh and some caution is worthwhile.
"One of my catch-phrases is that we decide what to believe by deciding who to believe."
LOL. I meet with a group of 5-10 diverse academics weekly. When I mentioned some covid related data, I got it horribly wrong for reasons unimportant here. Nobody corrected me. When I noticed my error I apologized and asked why they didn't point out my obvious error. The consensus was they don't look at the data and entirely tuned out my sharing of data. They look for someone they trust to form their opinion. They explicitly stated exactly what you propose.
Whatever real sociology is, your best posts are about it. To me that includes ChatGPT and other technological advancements, direction of academia and proposed changes, and yes most comments on links to consider. To me that does not include anti-woke stuff. Thank you
There will be the few who are capable of enhancing their own abilities by leveraging the use of these "AIs", while the many will simply see their own skills atrophy in reliance on, or being replaced by the "AIs".
Reading Legal Documents: Aside from drafting legal documents ChatGPT will read documents and abstract them and place the relevant data into models. The world of commercial real estate involves taking lots of leases and creating 10 year models from the lease contracts often by using industry standard model software like ARGUS. An AI interface would cut out lots of work done today by analysts. Loan documents could be read by ChatGPT and it could set up the billings. Finally, securitization of commercial real estate is often complicated by non-standard documents, but ChatGPT could provide a standard interpretation that fixes that problem.
‘Lately, I have been posting “Links to Consider” more frequently. In case you never click on those posts, you should know that I often add my own thoughts as comments to those links, so you are missing out on my thoughts as well as links that I find interesting. I hope that these posts are popular.”
I, for one, find this feature very useful. Please continue.
<blockquote>Alex and I agree that there is great potential use for it in education. It takes us quite a bit closer to Neal Stephenson’s vision in The Diamond Age of personalized instruction by computer. I predict that educators and institutions that choose to fight the technology rather than embrace it will go by the wayside.</blockquote>
I sincerely hope this doesn't happen. The problem with Stephenson's vision is that it eliminates the fixed, stable reference materials whose existence makes possible academic rigor. Instead the student winds up searching a vast body of untrustworthy data, much of it from anonymous sources or editors, which can easily be hijacked by selfish and corrupt people with elite access in order to deceive the masses and suppress dissent, just as has already happened to the Big "mainstream" Media. Without the ability to reliably check references the Scientific Method is no longer available.
I usually read all the articles you link to, especially if I want to comment on any specifically. There are other link aggregators I follow (Instapundit & Don Surber & MR plus others), but follow yours at a higher (90%+) hit rate than any others. Which is why I hang out here so much.
Was a good talk - I'd like to hear more about what folks are actually doing with AI, including chatGPT plus its competitors like Claude (?). Alex talked about Google glass coming back, but now I think it would be an ear-bud with a computerized voice reading the text message. Whether translated or sent as voicemail.
Otter.ai for voice to text will also be on my list to play with, more.
The intelligence will be simulated human intelligence, unless more programmed to be more forthcoming with estimates of accuracy for the facts. Connecting various ai bots together seems like a natural thing in the near future.
More verifiable accuracy to reduce hallucinations is important - as well as more clearly stating what worldview is being taken. Human belief has a huge good-bad pillar, and "neutral" is often really not an option.
Thanks for your thoughts on the direction you will take in the near future. I read your links to consider and follow those I find interesting in order to see other thoughts than the one or two you point out. So, I would like you to keep providing the information.
I found Monday's seminar stimulating. Alex and you provided a number of insights into where Chatgpt would take us. The most intriguing question came at the begining but didn't get much attention. "Just what kind of intelligence does Chatgpt possess?" I'd like to see more discussion of that.
I have given that some thought since Monday. I'll use music as an analogy for my thinking. My grandson, a musician, explained music to me as spacing and timing with the notes adding color. Obviously we need all three to have "music". Chatgpt already has tremendous abilities at spacing and timing (organizing its thoughts) but the notes come from humans. Its the collaboration of humans with Chatgpt that produce the music.
In that vein, I'd really get a kick out of a seminar where Chatgpt attended as a participant. I don't know how one would organize that. But, perhaps we should ask Chat to help with that.
Perhaps there will be an increasing role for micropayments for any email, or response to government as a way to filter out the abundance of potential spam.
Creating systems for verifying the authenticity of video or really anything will also be increasingly important. Perhaps I’m overrating the importance of recorded audio /video in establishing truth - technologies which are historically quite young - yet the notion of world in which there truly is no way to verify authenticity of video is concerning. We’re not prepared for such a world. Perhaps we just adopt a mindset where the default is no video can be believed. Alternatively in the same way that we have trusted sources for written facts, we adopt the same mentality with audio/video.
The counter is that we seem to get by now even with reasonable ability to selectively doctor/edit video.
Good discussion last night! I appreciated your perspective and Alex's.
Two suggestions (both of which you have probably already thought of):
1. It would be fun to get someone on who thinks that AI alignment is a big problem. Someone like Eliezer Yudkowsky? Or....Robin Hanson has debated Yudkowsky, maybe he would come on again to discuss the problem? Even though he (Hanson) doesn't seem to be worried about the problem, he would have the ability to steelman the side that does. He could give us both sides?
2. I would love to hear your take about the automation problem. More people seem worried that this technology is sufficiently different than previous technologies that it will cause significant employment problems. What's your best framework for thinking about this?
I read all of the “Links to Consider” content but unless I want to follow one of the links or share, I read it in my email. I’m guessing I’m not the only one doing that so you probably have a good bit of readership that the statistics aren’t capturing.
"It is fun to play with and there are many possible applications." (referring to ChatGPT)
Not sure if I disagree or simply misunderstand your comment yesterday about delving into WWW and ChatGPT. I agree with this one.,
As for Jaron’s claim about fakes, I'm skeptical it matters other than tangential questions such as pinching entry level jobs, flooding requests for comment, and use akin to plagarism in school. Or are those your concerns? My concerns regarding writing are whether it is clearly wrong factually and if it tells a compelling story. I don't think it's necessary to fine tune or expand on that to say my concerns are no different than for entirely human composition. "Fake" seems to be a misguided put down akin to luddite technophobia. But maybe that's too harsh and some caution is worthwhile.
"One of my catch-phrases is that we decide what to believe by deciding who to believe."
LOL. I meet with a group of 5-10 diverse academics weekly. When I mentioned some covid related data, I got it horribly wrong for reasons unimportant here. Nobody corrected me. When I noticed my error I apologized and asked why they didn't point out my obvious error. The consensus was they don't look at the data and entirely tuned out my sharing of data. They look for someone they trust to form their opinion. They explicitly stated exactly what you propose.
Whatever real sociology is, your best posts are about it. To me that includes ChatGPT and other technological advancements, direction of academia and proposed changes, and yes most comments on links to consider. To me that does not include anti-woke stuff. Thank you
There will be the few who are capable of enhancing their own abilities by leveraging the use of these "AIs", while the many will simply see their own skills atrophy in reliance on, or being replaced by the "AIs".
Reading Legal Documents: Aside from drafting legal documents ChatGPT will read documents and abstract them and place the relevant data into models. The world of commercial real estate involves taking lots of leases and creating 10 year models from the lease contracts often by using industry standard model software like ARGUS. An AI interface would cut out lots of work done today by analysts. Loan documents could be read by ChatGPT and it could set up the billings. Finally, securitization of commercial real estate is often complicated by non-standard documents, but ChatGPT could provide a standard interpretation that fixes that problem.
While I value the window on the anti-woke fringe that you provided me, I think this new direction is healthier. More creative, less reactionary.
‘Lately, I have been posting “Links to Consider” more frequently. In case you never click on those posts, you should know that I often add my own thoughts as comments to those links, so you are missing out on my thoughts as well as links that I find interesting. I hope that these posts are popular.”
I, for one, find this feature very useful. Please continue.
<blockquote>Alex and I agree that there is great potential use for it in education. It takes us quite a bit closer to Neal Stephenson’s vision in The Diamond Age of personalized instruction by computer. I predict that educators and institutions that choose to fight the technology rather than embrace it will go by the wayside.</blockquote>
I sincerely hope this doesn't happen. The problem with Stephenson's vision is that it eliminates the fixed, stable reference materials whose existence makes possible academic rigor. Instead the student winds up searching a vast body of untrustworthy data, much of it from anonymous sources or editors, which can easily be hijacked by selfish and corrupt people with elite access in order to deceive the masses and suppress dissent, just as has already happened to the Big "mainstream" Media. Without the ability to reliably check references the Scientific Method is no longer available.
The Lanier essay appears to have been taken down. Fortunately, Google cached it.
https://webcache.googleusercontent.com/search?q=cache:g7vmXO7RXoEJ:https://www.tabletmag.com/sections/news/articles/oy-ai-jaron-lanier&cd=1&hl=en&ct=clnk&gl=us
I usually read all the articles you link to, especially if I want to comment on any specifically. There are other link aggregators I follow (Instapundit & Don Surber & MR plus others), but follow yours at a higher (90%+) hit rate than any others. Which is why I hang out here so much.
Was a good talk - I'd like to hear more about what folks are actually doing with AI, including chatGPT plus its competitors like Claude (?). Alex talked about Google glass coming back, but now I think it would be an ear-bud with a computerized voice reading the text message. Whether translated or sent as voicemail.
Otter.ai for voice to text will also be on my list to play with, more.
The intelligence will be simulated human intelligence, unless more programmed to be more forthcoming with estimates of accuracy for the facts. Connecting various ai bots together seems like a natural thing in the near future.
More verifiable accuracy to reduce hallucinations is important - as well as more clearly stating what worldview is being taken. Human belief has a huge good-bad pillar, and "neutral" is often really not an option.
Thanks for your thoughts on the direction you will take in the near future. I read your links to consider and follow those I find interesting in order to see other thoughts than the one or two you point out. So, I would like you to keep providing the information.
I found Monday's seminar stimulating. Alex and you provided a number of insights into where Chatgpt would take us. The most intriguing question came at the begining but didn't get much attention. "Just what kind of intelligence does Chatgpt possess?" I'd like to see more discussion of that.
I have given that some thought since Monday. I'll use music as an analogy for my thinking. My grandson, a musician, explained music to me as spacing and timing with the notes adding color. Obviously we need all three to have "music". Chatgpt already has tremendous abilities at spacing and timing (organizing its thoughts) but the notes come from humans. Its the collaboration of humans with Chatgpt that produce the music.
In that vein, I'd really get a kick out of a seminar where Chatgpt attended as a participant. I don't know how one would organize that. But, perhaps we should ask Chat to help with that.
I think that getting bots to chat is a great idea. But I don't think it'll eliminate hallucination, because none of them have access to ground truth. I explain at some length, https://new-savanna.blogspot.com/2023/01/theres-truth-lies-and-theres-chatgpt.html
Perhaps there will be an increasing role for micropayments for any email, or response to government as a way to filter out the abundance of potential spam.
Creating systems for verifying the authenticity of video or really anything will also be increasingly important. Perhaps I’m overrating the importance of recorded audio /video in establishing truth - technologies which are historically quite young - yet the notion of world in which there truly is no way to verify authenticity of video is concerning. We’re not prepared for such a world. Perhaps we just adopt a mindset where the default is no video can be believed. Alternatively in the same way that we have trusted sources for written facts, we adopt the same mentality with audio/video.
The counter is that we seem to get by now even with reasonable ability to selectively doctor/edit video.
Good discussion last night! I appreciated your perspective and Alex's.
Two suggestions (both of which you have probably already thought of):
1. It would be fun to get someone on who thinks that AI alignment is a big problem. Someone like Eliezer Yudkowsky? Or....Robin Hanson has debated Yudkowsky, maybe he would come on again to discuss the problem? Even though he (Hanson) doesn't seem to be worried about the problem, he would have the ability to steelman the side that does. He could give us both sides?
2. I would love to hear your take about the automation problem. More people seem worried that this technology is sufficiently different than previous technologies that it will cause significant employment problems. What's your best framework for thinking about this?
Thanks!
I read all of the “Links to Consider” content but unless I want to follow one of the links or share, I read it in my email. I’m guessing I’m not the only one doing that so you probably have a good bit of readership that the statistics aren’t capturing.