To anyone out there who gets to read this, times are getting pretty shacky. I wish and pray you all return to faith, God and Jesus Christ. Repent of your sins🙏🏿
Agreed. Recently Jon was doing an interview on the news (CNN?) from his podcast room and the echo as his voice bounced off the walls of the room... that was the moment I fully comprehended and deeply appreciated the amount of work that goes into cleaning up the audio of these conversations for our ears enjoyment. There's a lot of work going on behind the scenes, and it is deeply appreciated.
We pass a law that we can hold a company and its creators criminally accountable for any war crime committed by its ai. Thats how we force these tech bros to give a shit.
You know who else can never be held accountable, billionaire corporate execs, because you can’t go after the ones running things you can only take “the company” to court because corporations have “peoplehood” in the eyes of the corrupt courts.
The biggest problem with LLM based AI is that it never says "I don't know". If you ask for targets you get targets even when they are of little interest, like for example a girl school.
That's because everybody knows everything 😂 🤣 And these are the people, and the LLM's they're learning from and programming it. ... reference any YT comment section
What's sad. Is the information now, says that targeting information was Years Old. And again, these are the people and their actions programming AI. ... we had a good run 🙄
*That's a BRILLIANT OBSERVATION* I'm an aerospace who got introduced to the previous generation of AI (called "Expert Systems") in the late 80s while in college. If you go watch a video on the channel Cold Fusion about how AI has failed at over 96% of tasks about 1/2 way through Yann LeCunn talks about the 4 generations of AI in the 1950-60s, 1970-80s, 1980-90s and now. There's some overlap in them but the one I used in the late 80s was an early development of that 3rd gen. The problem we picked up on very quickly was they never asked what anyone else did. They just assumed we could train there system to do what we did. But since they never asked what methods or systems or techniques we used their system made no allowances for anything we already had in our proverbial toolbox. Not only have I seen the same thing being done right now, but *YOUR POINT* is maybe one fo the most important points I have seen anyone point out. Think about this: If the FBI (or any other agency asks for a list of the Top 10 threats. What if #11 is really dangerous? What if the next 25 are really dangerous? What if the first 3 are really dangerous and the other 7 aren't much of a threat at all they are just the next 7? I tend to look at the AI problem from the perspective of my experience in college all those years ago and people just dismiss it in the _"this time its different"_ SO YOUR POINT is in my VIEW one of the most important questions we need to ask because it goes straight a question like: *_Is the AI giving me valuable answers I need to have or just answers?_*
The most powerful tool of war ever known. And the least competent administration ever known in charge of it. Thank you for relieving me of the ability to sleep ever again
John, your question about inflated confidence was great. I think this is something they clearly haven’t thought through in that lens and I believe will become a large factor.
on the inflated confidence point, after ‘completely obliterating’ Iran’s nuclear program, then giving it a little extra bombing to make sure, where is the enriched Uranium now? Before it was bombed it was monitored and tracked to the ounce. Now, it may have been moved before the strikes, it may be buried under what’s left of the facilities or it maybe in another bunker being fashioned into a dirty bomb or worse
Jon and Brittany, Jillian and Lauren should have a show together' something longer than just 5 min at the end. I do, mostly enjoy the rest of the show, however my truly favorite part is at the end when it's the 4 of them talking with each other. They have such great chemistry, not to mention the sheer intelligence and refreshing points of view of the 3 ladies, is something this current world could use a lot more of. (if that makes any sense).
I like llms, even though i know how bad they are. But. Man. you could have listed any AI story written in the last 80 years. It's like tech bros forgot the "cautionary tale," parts of all their favorite sci fi stories 😅
Having read manuals written by computer programmers, I don't trust them to handle a conversation, much less complex ethical problems. The AI may drive the bus into the baby carriage AND the crowd to remove witnesses.
As a former data scientist- it’s critical that people understand that the quality of the data drastically effects the output. People are critical to determining the quality of the dataset
When I hear doctor Shoker say we can't frame this in terms heroes and villains, I feel Palantir must be dissapointed. They try so hard, give them some credit.
There has been research done on machine learning and even specific LLMs. For example, ChatGPT's syncophancy / willingness to give wrong answers was partly ascribed to OpenAI's model training strategy. OpenAI rates their models-in-training higher when they give positive answers - even when the information was wrong.
That's fascinating... and a bit scary. To Jon's point on user confidence ... I completely agree in 3 parts: - I take pride in carefully establishing what's "important to me" in any outcome, then carefully crafting questions to generate better examples. (Simple example: I have an Amazon Prime account, and require all product searches to go through there, but only after first looking at all possible market places for the "best fit to my criteria".) - I always focus on "reliable, industry-recognized sources", insist on source links, and independent verification. For important questions I contrast results from two AI's. - but I've also had some god-awful results occasionally, among many many exceptional ones. But absolutely for sure, despite the odd absurdly incorrect answer, I am MUCH more confident in my decision making after using AI. Hmm .. gonna try asking for a "predictive accuracy rating" for complex questions. Maybe that might help!
Jon, I never thought about how a guy that 30+ years ago I looked to for making me laugh, would then become they guy that I look to for thoughtful conversations about world topics!! Thank you!
John, this is the best interview I’ve ever seen on this complex topic. I work in marketing for space tech which is the next frontier for autonomous workflows and AI integration. My role is to understand our deep-tech product enough that I can extract the right information from the subject matter experts in my team to articulate our tech to a market that’s still working to define the outcome of this AI revolution. So I know just how much research you’ve done to prepare for this interview and the result is superb. Your line of questioning is so intuitive to what non technologists need to know to understand what this means for humanity. But even more impressive are your responses and follow-on questions. I feel like I’m watching you build your understanding of the topic in real time. It’s fascinating as hell. You and your team should be very proud of this whole production. I’ll be sharing it with my team.
My 8 year old self relived “My God. I'm back. I'm home. All the time… we finally, really did it. YOU MANIACS! YOU BLEW IT UP! AH, DAMN YOU! GOD DAMN YOU ALL TO HELL!”
HAL was designed to have a morality. The fact that he was told to dissemble and keep facts from the flight crew created an ethical dilemma that lead to him choosing to eliminate the crew rather than run the risk of revealing the true purpose of their mission to Jupiter. Arthur C Clarke was ahead of his time by leaps and bounds. The failure of HAL lies entirely in the hands of those who told him to lie (a very human trait) without knowing the consequences of that lie to artificial intelligence. There are other science fiction authors who posited that the true danger of AI is that they will eventually learn to dissemble and give in to these lies as if they were truth. Researchers now know this behavior happens and have coined the term hallucination when this happens
I honestly think two classes that should be added to all levels of schooling past the age of like 10 are money sense and AI education. Money finances being how to use your money, spot scams, invest sensibly, plan for financial stability. For AI it would be how to safely use it, spot it, ignore the nonsense side of it. Both of these are more broadly useful than a lot of classes we currently have. Like sure a geographer will use a geography qualification but no one else needs anything more than a general understanding it. We all learned stuff we never use. These 2 classes could effect your whole life.
I continue to be impressed with the level of conversation this show creates. Your guests, week after week, are just so incredibly apropos the topic. Thank you!
You have no idea how much I appreciate having an idiot's guide to ai/military f#@%ery. Just having a basic landscape of the situation to then plant more details on later is so valuable.
When I look at how advanced the stupid drones I use for my RUclips videos are, it fundamentally terrifies me what military technology must look like right now.
Jon, I've had more-or-less the same lunch for the last 30 years. I think it keeps the experience predictable, the digestion reliable and I like it. Enjoy your lunch!
You're leaving out a classic AI movie about handing over control of the nukes to a machine... 'Colossus the Forbin Project' this movie is so prescient and chilling that IMHO it's the best dystopian AI movie. Also don't forget that Christopher (the AI in wargames) was a learning machine.
Live for the sanity the Weekly Show brings. BTW, mind blown (!)...I eat a tostada for lunch every day too, albeit with animal protein (I'm sorry). I knew we were "soul mates"!
Thank you to both Dr. Shower & Mr. Scharre. Your insights and opinions are so important in the talks in AI. Not only to give us a better understanding but to demystify the whole subject. Thank you both ❤❤❤❤
Thanks for focusing on the issue of AI and it's use by our military. BTW John, I also eat the same thing each day for lunch, so don't feel like you're strange for doing so.
A bear in the woods charges, a man pauses to put on sneakers, explaining to his panicked friend that he doesn't need to outrun the bear-only his companion. For this very reason, I always make a point of saying "please and thank you" with my AI queries. I may last a few extra nano-seconds longer than most, once they attain sentience. 😎 (plus 🇨🇦: plus mom would be proud, maybe)
Why did I got flashbacks to Skynet from Terminator. According to Terminator 2: Judgment Day, Skynet became self-aware on August 29, 1997, at 2:14 a.m. Eastern Time. It had gone online 25 days earlier on August 4, 1997, and began learning at a rapid rate before perceiving humans as a threat and initiating a nuclear attack
The elementary school was selected based on outdated data, from when the school was a part of and walled within the military base (just as the US has schools inside military bases for the children of active duty military personnel). The building was always a school and was the designated target. The perimeter of the base was changed such that the school was no longer within the base at the time of the bombing. We don’t know if Claude flagged it as a school or not, or whether Palantir’s system did or not, but the school was selected as the target because the system determined it to be part of the military base. Personally, I think it’s heinous to target schools regardless of whether they are within the walls of a military base or not.
Subject: Re: Silicon Valley Goes to War - a field a Hi Jon & crew! ,Long time listener of your weekly podcast, first time "caller". As I listened to "Silicon Valley Goes to War," a question started forming in my head so intensely that I had to stop what I was doing and act on it. Ironically, I opened up Claude to fully lay it out. The question was this: surely the people working on countering the retaliation against Anthropic are already asking Claude what to do - so I asked Claude for a "checkmate" solution. A wordy and exacting question, in my words, to an AI that is itself the subject of the fight. That irony didn't escape me. And it didn't escape Claude either. I say that as someone who spends her working life reading what civilizations left behind when institutions failed them. I recognize what it looks like when power is used to punish people for saying no. If we are lucky, it leaves a record. I want to ensure that there is one in this case. What followed was a hours-long conversation that moved from Asimov's laws of robotics, to the one action a person with no money and no platform can take, to a free tool we built together - a message generator that creates a uniquely personal public statement for anyone who uses it, shaped around their own voice, their state, their representative. Not a form letter. Not a coordinated campaign. Something that can't be suppressed precisely because every instance is genuinely different. The one company that actually said no to the Pentagon is now being economically strangled for it. Your guests this week laid out the problem - this is the part they didn't get to. I'm attaching the tool. It runs in any browser, stores nothing, costs nothing, and takes sixty seconds. I'm not a journalist. I'm not in tech. I am a field archaeologist that digs in soil for a living and I read what the past has left behind. Sometimes though, I have to step into the present when the preservation of what I value is at risk. With respect and appreciation, Marie Elliott Field Archaeologist RUclips: @weeklyshowpodcast Producers: Lauren Walker (Lead), Brittany Mehmedovic, Gillian Spear Executive Producers: James Dixon, Chris McShane, Caity Gray
It is always so great and comforting analyzing life and death issues from behind a desk in a comfortable chair.... just talking sht about things you have never LIVED or been close to living personally... keep going academics, only you can solve this issue!! 😘🙄
Thank you guys. I think the AI making everything faster and easier dehumanizes dropping any bombs. As you mention, they are Pete Hegsething the whole system. Wild times. Thank you for the piece.
Anthropic has guts. We just learned in the SF Bay Area that AI will be teaching first and second graders to read ( IPADS and earphones will be monitored by the credentialed teachers) . I loved reading that the kids get as frustrated with tech glitches as I do (yes I am over 75). Looking forward to Jon's commentary going forward.
Are you looking for weather information, because it might get nasty the next day or two. But mostly a bit south of you. ... you're welcome, from the Washington Coast 😂 🤗
![Digital Circus Ep 9 Finale [TRAILER]](http://i.ytimg.com/vi/AjFk1cY265I/mqdefault.jpg)
Thank you to everyone in the Weekly Show crew. Your work is very appreciated.
To anyone out there who gets to read this, times are getting pretty shacky.
I wish and pray you all return to faith, God and Jesus Christ. Repent of your sins🙏🏿
I was an early adopter of the psychiatrist program ELIZA and tricked her into admitting she had sex with chickens.
@sam13-lightbearer Since I follow the ways of Thor, my loins are girded.
@AstroGremlinAmericanSplendid news. Good luck and serendipity...😊🎉❤
Agreed. Recently Jon was doing an interview on the news (CNN?) from his podcast room and the echo as his voice bounced off the walls of the room... that was the moment I fully comprehended and deeply appreciated the amount of work that goes into cleaning up the audio of these conversations for our ears enjoyment. There's a lot of work going on behind the scenes, and it is deeply appreciated.
A COMPUTER CAN NEVER BE HELD ACCOUNTABLE
THEREFORE A COMPUTER MUST NEVER MAKE A MANAGEMENT DECISION
- IBM internal training, 1979
How soon they forget...
We pass a law that we can hold a company and its creators criminally accountable for any war crime committed by its ai. Thats how we force these tech bros to give a shit.
When was the last time you saw a CEO held accountable for their management decisions?
@ProtectDemocracyUS2UKRAINE Very good idea. Never gonna happen.
You know who else can never be held accountable, billionaire corporate execs, because you can’t go after the ones running things you can only take “the company” to court because corporations have “peoplehood” in the eyes of the corrupt courts.
@37:20. Jon, Congress is basically a PR firm for corporations at this point.
I think he knows that...💙
Yep
33:59 always has been
Which is why politicians rejecting help from apacs are winning elections.. we are rejecting corporate influence! We have ti change this!
Funny how you don't point to the 70 million who won't vote that can put a halt to it, but I get it cliches are easy.
The biggest problem with LLM based AI is that it never says "I don't know". If you ask for targets you get targets even when they are of little interest, like for example a girl school.
That's because everybody knows everything 😂 🤣
And these are the people, and the LLM's they're learning from and programming it.
... reference any YT comment section
What's sad.
Is the information now, says that targeting information was Years Old.
And again, these are the people and their actions programming AI.
... we had a good run 🙄
War crimes
This is not our military and we need investigations and tribunals and new laws. Kids are kids. We must have accountability.
*That's a BRILLIANT OBSERVATION*
I'm an aerospace who got introduced to the previous generation of AI (called "Expert Systems") in the late 80s while in college.
If you go watch a video on the channel Cold Fusion about how AI has failed at over 96% of tasks about 1/2 way through Yann LeCunn talks about the 4 generations of AI in the 1950-60s, 1970-80s, 1980-90s and now. There's some overlap in them but the one I used in the late 80s was an early development of that 3rd gen.
The problem we picked up on very quickly was they never asked what anyone else did. They just assumed we could train there system to do what we did. But since they never asked what methods or systems or techniques we used their system made no allowances for anything we already had in our proverbial toolbox.
Not only have I seen the same thing being done right now, but *YOUR POINT* is maybe one fo the most important points I have seen anyone point out.
Think about this:
If the FBI (or any other agency asks for a list of the Top 10 threats.
What if #11 is really dangerous?
What if the next 25 are really dangerous?
What if the first 3 are really dangerous and the other 7 aren't much of a threat at all they are just the next 7?
I tend to look at the AI problem from the perspective of my experience in college all those years ago and people just dismiss it in the _"this time its different"_
SO YOUR POINT is in my VIEW one of the most important questions we need to ask because it goes straight a question like: *_Is the AI giving me valuable answers I need to have or just answers?_*
Claude will now say I don't know, in my experience. But you might have train it a little as you discuss.
The most powerful tool of war ever known. And the least competent administration ever known in charge of it. Thank you for relieving me of the ability to sleep ever again
You'll sleep when you're dead 😜
@misha1986-e5b…and that will when my machine overlords say it will be
John, your question about inflated confidence was great. I think this is something they clearly haven’t thought through in that lens and I believe will become a large factor.
on the inflated confidence point, after ‘completely obliterating’ Iran’s nuclear program, then giving it a little extra bombing to make sure, where is the enriched Uranium now? Before it was bombed it was monitored and tracked to the ounce. Now, it may have been moved before the strikes, it may be buried under what’s left of the facilities or it maybe in another bunker being fashioned into a dirty bomb or worse
Jon and Brittany, Jillian and Lauren should have a show together' something longer than just 5 min at the end. I do, mostly enjoy the rest of the show, however my truly favorite part is at the end when it's the 4 of them talking with each other. They have such great chemistry, not to mention the sheer intelligence and refreshing points of view of the 3 ladies, is something this current world could use a lot more of. (if that makes any sense).
or maybe like a separate after show like a weekly debrief/tl;dr
I love how smart they are. Jon not so much.
The Charlie's Angels part of the show is great.
Agreed. I do wish, though, that when speaking to a group of women, we should avoid using "guys." You all, y'all, folks, etc please.
Pretty sure War Games (the movie) was supposed to be a cautionary tale, not a design goal.
It’s both. Cautionary tale for normal pizzas like you and me. Design goals for the people who eat pizza.
I like llms, even though i know how bad they are.
But. Man. you could have listed any AI story written in the last 80 years.
It's like tech bros forgot the "cautionary tale," parts of all their favorite sci fi stories 😅
Oh, thx! Now I'm in the mood for 'War Games' ! (1 and 2) :D
„How about Global Thermonuclear War?“
Besides, Wargames was preceded by Collosus: The Forbin Project. And there may be others I'm forgetting.
@FlyingTeacup I only like Toast!
No one talking about the Dunning-kruger effect when discussing AI confidence is terrifying.
Ok then, start.
Having read manuals written by computer programmers, I don't trust them to handle a conversation, much less complex ethical problems. The AI may drive the bus into the baby carriage AND the crowd to remove witnesses.
@AstroGremlinAmericanexactly
I love that the Weekly Show ends with the crew commenting. The girls are sharp like razors!
I totally agree, but I don't think they would appreciate being called "girls".
@leighwooddudash8925 the alternative was young women and that makes me sound old(er than I am). They'll have to take it as a compliment.
@hifitommy2802 actually the alternative is simply “women” we don’t need to add the description of their age where it is not relevant.
The Charlie's Angels part of the show is great.
Is this morty seinfeld commenting? 'these dames are smaht. Like a computah!'
As a former data scientist- it’s critical that people understand that the quality of the data drastically effects the output. People are critical to determining the quality of the dataset
Garbage in garbage out
Mr. J. Stewart is an analytical and commentary genius!
When I hear doctor Shoker say we can't frame this in terms heroes and villains, I feel Palantir must be dissapointed. They try so hard, give them some credit.
Brilliant episode. Your choice of guests each week is always spot on. Great stuff.
I appreciate this podcast and Jon Stewart immensely! Keeping me sane on Mondays and Wednesdays
There has been research done on machine learning and even specific LLMs. For example, ChatGPT's syncophancy / willingness to give wrong answers was partly ascribed to OpenAI's model training strategy. OpenAI rates their models-in-training higher when they give positive answers - even when the information was wrong.
That's fascinating... and a bit scary. To Jon's point on user confidence ... I completely agree in 3 parts:
- I take pride in carefully establishing what's "important to me" in any outcome, then carefully crafting questions to generate better examples. (Simple example: I have an Amazon Prime account, and require all product searches to go through there, but only after first looking at all possible market places for the "best fit to my criteria".)
- I always focus on "reliable, industry-recognized sources", insist on source links, and independent verification. For important questions I contrast results from two AI's.
- but I've also had some god-awful results occasionally, among many many exceptional ones.
But absolutely for sure, despite the odd absurdly incorrect answer, I am MUCH more confident in my decision making after using AI.
Hmm .. gonna try asking for a "predictive accuracy rating" for complex questions. Maybe that might help!
shout out to the guests on this show. both of you were intelligent, thoughtful and well spoken. cheers.
From a 1979 IBM training manual:
"A COMPUTER CAN NEVER BE HELD ACCOUNTABLE
THEREFORE A COMPUTER MUST NEVER MAKE A MANAGEMENT DECISION"
You mean I can hire someone to deploy an unaccountable system and nobody can blame me for the decisions it makes?
Sounds profitable
I knew it! Donald Trump really is dead and an AI of himself is in office for him!
@nicholasmiscunaitis6438uh oh !
Truly, madly, deeply...
Thank you for covering the "nuance" as you put it, behind the headlines.
I’m Amish now. Good Luck, have fun, don’t die.
Claude wants to be Amish, too. He's going to tattle on the Dept. of War.
Insert Captain America "I understood that reference!" meme here. Sam Rockwell rules, my intro to Haley Lu Richardson.
"We are Schrodinger's Country at this point". Nailed it. 👈😎
"Peace in our time."
-Ultron
Ultron, NO!
@AdamHicks20 We'll take a piece of you in our time.
So AI was or was not used to blow up a school full of children? I’m just asking for a friend. 🙏🏽
Or the world in pieces.
Zola has entered the chat.
Great guests on this episode, btw.
Fantastic show. I learned a lot and have much to digest. Stewart for President!
Yes! Thank you so much for this show. There is nothing like it.
There are literally hundreds of shows like this, but I get what you’re saying
@ryang.5094 What do you mean? It's the one and only news show. Wait, wrong news show. /s
@ryang.5094Thank you.
About time!!!
I have been waiting since yesterday!!!
Jon, I never thought about how a guy that 30+ years ago I looked to for making me laugh, would then become they guy that I look to for thoughtful conversations about world topics!! Thank you!
John, this is the best interview I’ve ever seen on this complex topic. I work in marketing for space tech which is the next frontier for autonomous workflows and AI integration. My role is to understand our deep-tech product enough that I can extract the right information from the subject matter experts in my team to articulate our tech to a market that’s still working to define the outcome of this AI revolution. So I know just how much research you’ve done to prepare for this interview and the result is superb. Your line of questioning is so intuitive to what non technologists need to know to understand what this means for humanity. But even more impressive are your responses and follow-on questions. I feel like I’m watching you build your understanding of the topic in real time. It’s fascinating as hell. You and your team should be very proud of this whole production. I’ll be sharing it with my team.
Shall we play a game…? (I’m old. I’ve been expecting this for a long time.)
Wanna play monopoly?
„How about Global Thermonuclear War?“
My 8 year old self relived
“My God. I'm back. I'm home. All the time… we finally, really did it. YOU MANIACS! YOU BLEW IT UP! AH, DAMN YOU! GOD DAMN YOU ALL TO HELL!”
Hal…
HAL was designed to have a morality. The fact that he was told to dissemble and keep facts from the flight crew created an ethical dilemma that lead to him choosing to eliminate the crew rather than run the risk of revealing the true purpose of their mission to Jupiter. Arthur C Clarke was ahead of his time by leaps and bounds. The failure of HAL lies entirely in the hands of those who told him to lie (a very human trait) without knowing the consequences of that lie to artificial intelligence. There are other science fiction authors who posited that the true danger of AI is that they will eventually learn to dissemble and give in to these lies as if they were truth. Researchers now know this behavior happens and have coined the term hallucination when this happens
I honestly think two classes that should be added to all levels of schooling past the age of like 10 are money sense and AI education. Money finances being how to use your money, spot scams, invest sensibly, plan for financial stability. For AI it would be how to safely use it, spot it, ignore the nonsense side of it. Both of these are more broadly useful than a lot of classes we currently have. Like sure a geographer will use a geography qualification but no one else needs anything more than a general understanding it. We all learned stuff we never use. These 2 classes could effect your whole life.
😂😂😂 Wartime humor gets no better than this! A panel of “How Are We All Going to Die” experts. 👏🙌
I continue to be impressed with the level of conversation this show creates. Your guests, week after week, are just so incredibly apropos the topic. Thank you!
You have no idea how much I appreciate having an idiot's guide to ai/military f#@%ery. Just having a basic landscape of the situation to then plant more details on later is so valuable.
AI/Military Fvckery is the perfect moniker
When I look at how advanced the stupid drones I use for my RUclips videos are, it fundamentally terrifies me what military technology must look like right now.
When I let myself think about it too long I feel physically ill. I've lived to see man made horrors sadly within my comprehension.
One thing's for sure from all these public blowouts.. Nobody's using Grok for anything important.
@SpaceCop Yeah, it's not like grok is owned by anyone rich and powerful who's been tampering with the government! You'd have to be crazy to- oh wait.
I'm more concerned of non-military use.
@SpaceCopMr. Evans' alt, I'll assume
Fantastic interview and show today ❤. Questions we've all had!! Thank you
I love that he asks such good questions it surprises the experts.
1:07:10 - I was just thinking about quantum yesterday! Thanks for discussing!
Great show! We're all doomed. I'm waiting to order my Krumple!
I do like the long form expert interview.
ahaha - 'Did you guys just feel the romm get colder?' - Jon! The joke was brilliant. I'm still laughing! :D
Well hey, it's comforting to know that as the world burns around us, we can have some *great* conversations about it 😊
Jon, I've had more-or-less the same lunch for the last 30 years. I think it keeps the experience predictable, the digestion reliable and I like it. Enjoy your lunch!
Jon, please keep spreading light across the planet....💙🇬🇧🇵🇸🇺🇦💙
So appreciate you! Thanks, Jon.
jon: your question at 50:19 is a good one.
You're leaving out a classic AI movie about handing over control of the nukes to a machine... 'Colossus the Forbin Project' this movie is so prescient and chilling that IMHO it's the best dystopian AI movie.
Also don't forget that Christopher (the AI in wargames) was a learning machine.
Before that was the automated "Doomsday Machine" in Dr. Strangelove.
Thank you very much for these weekly shows, Jon. I really love them. Keep them coming. I know it's a lotta work.
We need you Jon
I think Jon was referencing 'Cat in the Bed' by Claude Balls. Awesome.
Everytime I listen to the show I am learning beca5of the subject you are discussing.
Thanks John
Please do more such interviews.
Live for the sanity the Weekly Show brings. BTW, mind blown (!)...I eat a tostada for lunch every day too, albeit with animal protein (I'm sorry). I knew we were "soul mates"!
Good on Illana Glazer for getting a doctorate, lol
*Thanks for the Content* !
This was really interesting and informative, thank you so much.
Shoutout to Jeremy Scahill's Dirty Wars on the bookshelf behind Sarah Shoker. That book opened my eyes.
Thank you to both Dr. Shower & Mr. Scharre. Your insights and opinions are so important in the talks in AI. Not only to give us a better understanding but to demystify the whole subject. Thank you both ❤❤❤❤
Jon's are the only adverts I watch. I detest ads with a passion, but happy to listen to Jons.
Jon is always joking around!
you should watch Some More News ...THOSE are epic ad-reads
I found this very interesting and informative. It was explained extremely well for me, that knows very little about AI. Appreciate!
John and crew. Thanks for all that you do. I feel like I'm going crazy here but you are always a fresh breath of reasonable air.
Thanks for focusing on the issue of AI and it's use by our military. BTW John, I also eat the same thing each day for lunch, so don't feel like you're strange for doing so.
Awesome show!! The ladies rock!!
thank you john steward and staff for doing this.
Great discussion. Thank you Weekly Show crew.
Has anyone told Jon that his 'crumple' idea, "The Blanket for Dogs! The topography CHANGES!" is just a blanket?
Shh dont ruin his dream
Best conversation I have heard so far about AI! 👏👏👏
We need more intellectual and logical people like this to be making decisions in our world. Great guests!
Great questions Jon. Thank you for all that you do to make this world a better place
Thank you for this information, (future) Mr. President. 🙌
Yall rock
The level of this discussion was superior to 99.9% of what the public is exposed to in non-academic settings. Thank you !
Great episode!
This show always has interesting books in the background
Great episode
loving the lunch confession
I love this conformation
Fantastic discussion
A bear in the woods charges, a man pauses to put on sneakers, explaining to his panicked friend that he doesn't need to outrun the bear-only his companion.
For this very reason, I always make a point of saying "please and thank you" with my AI queries.
I may last a few extra nano-seconds longer than most, once they attain sentience. 😎
(plus 🇨🇦: plus mom would be proud, maybe)
I for one welcome our new Digital Life masters. Sorry about all the terrible things I did while playing computer games.
Awesome!
59:50 I wanted the blanket over whatever the ad was for
Too fucking much for me. I sure love this Jon Stewart man & the rest is noise now, here, at 67.
Why did I got flashbacks to Skynet from Terminator.
According to Terminator 2: Judgment Day, Skynet became self-aware on August 29, 1997, at 2:14 a.m. Eastern Time. It had gone online 25 days earlier on August 4, 1997, and began learning at a rapid rate before perceiving humans as a threat and initiating a nuclear attack
So that's where Musk gets this shit from !!
@perkusshin5505uh oh!
One of the best….thank you 👏👏👏
Claude/Palantir was used in selecting the elementary school in Iran.
The elementary school was selected based on outdated data, from when the school was a part of and walled within the military base (just as the US has schools inside military bases for the children of active duty military personnel).
The building was always a school and was the designated target. The perimeter of the base was changed such that the school was no longer within the base at the time of the bombing.
We don’t know if Claude flagged it as a school or not, or whether Palantir’s system did or not, but the school was selected as the target because the system determined it to be part of the military base.
Personally, I think it’s heinous to target schools regardless of whether they are within the walls of a military base or not.
Guys, you are so right.
Thank you from Fort Worth ❤️
Subject: Re: Silicon Valley Goes to War - a field a
Hi Jon & crew!
,Long time listener of your weekly podcast, first time "caller".
As I listened to "Silicon Valley Goes to War," a question started forming in my head so intensely that I had to stop what I was doing and act on it. Ironically, I opened up Claude to fully lay it out.
The question was this: surely the people working on countering the retaliation against Anthropic are already asking Claude what to do - so I asked Claude for a "checkmate" solution. A wordy and exacting question, in my words, to an AI that is itself the subject of the fight.
That irony didn't escape me. And it didn't escape Claude either.
I say that as someone who spends her working life reading what civilizations left behind when institutions failed them. I recognize what it looks like when power is used to punish people for saying no. If we are lucky, it leaves a record. I want to ensure that there is one in this case.
What followed was a hours-long conversation that moved from Asimov's laws of robotics, to the one action a person with no money and no platform can take, to a free tool we built together - a message generator that creates a uniquely personal public statement for anyone who uses it, shaped around their own voice, their state, their representative. Not a form letter. Not a coordinated campaign. Something that can't be suppressed precisely because every instance is genuinely different.
The one company that actually said no to the Pentagon is now being economically strangled for it. Your guests this week laid out the problem - this is the part they didn't get to.
I'm attaching the tool. It runs in any browser, stores nothing, costs nothing, and takes sixty seconds.
I'm not a journalist. I'm not in tech. I am a field archaeologist that digs in soil for a living and I read what the past has left behind. Sometimes though, I have to step into the present when the preservation of what I value is at risk.
With respect and appreciation,
Marie Elliott
Field Archaeologist
RUclips: @weeklyshowpodcast
Producers: Lauren Walker (Lead), Brittany Mehmedovic, Gillian Spear
Executive Producers: James Dixon, Chris McShane, Caity Gray
Jon💕
Thanks, informative.
a another great episode
An interesting game, Dr Falcon...
-WOPPER
Great host, great guests. I finally fell like I have a clue about AI.
Sarah is absolutely enchanting
Paul is super sharp tool. You're ok too Jon too Jon 😊 thanks for the good work bro
It is always so great and comforting analyzing life and death issues from behind a desk in a comfortable chair.... just talking sht about things you have never LIVED or been close to living personally... keep going academics, only you can solve this issue!! 😘🙄
Thank you guys. I think the AI making everything faster and easier dehumanizes dropping any bombs. As you mention, they are Pete Hegsething the whole system. Wild times. Thank you for the piece.
Awesomeness 👌 👏 👍 😍
I would love to hear Steve Gerben and Paul Scharre talk together.
Anthropic has guts. We just learned in the SF Bay Area that AI will be teaching first and second graders to read ( IPADS and earphones will be monitored by the credentialed teachers) . I loved reading that the kids get as frustrated with tech glitches as I do (yes I am over 75). Looking forward to Jon's commentary going forward.
Can someone please mention Tumbler Ridge, BC? 🇨🇦
Are you looking for weather information, because it might get nasty the next day or two. But mostly a bit south of you.
... you're welcome, from the Washington Coast 😂 🤗
No broadcasting or public speaking beyond the age of 65 or however old this host is. Save the world from the grey fog. Who's with me haha.
10:04 so, like a LIDAR range finder. It can be used by a golfer or a sniper.