The number you have dialed is not in service.
Please check the number and dial again, or dial 611 for customer assistance.
Message 53, P-A-O-3.
The number you have dialed is not in service.
Thank you.
Thank you.
Thank you.
Thank you.
I'm back. It's nice to be back, and it's nice to be home for a little while.
Well, it's nice to have you here. So, welcome.
I appreciate that.
Welcome back.
Hey, you know, you have kids.
And I'm wondering, are the kids of today fascinated by the space program?
Because, you know, for the first time in about 50 years, we're actually embarking on missions
to the moon. And I remember when I was a kid the first time, wow, that was really fascinating.
Now I'm a kid the second time, and it's really still fascinating to me.
But I'm wondering, how is it with the younger generation?
Or has it not come up at all?
Oh, it's come up quite a bit. And my oldest son is really quite fascinated with avionics and flight systems and planes and, you know, the entire aerospace industry.
So, no, we watched this live. We watched the Artemis II launch. We were watching it again as they went around the moon, through the dark side of the moon, and lost communications for, you know, approximately 41 minutes.
This whole thing is intensely fascinating for him.
And I think it's one of the very few things about which we can remain to be proud as Americans.
So I think that's pretty cool for kids to see.
Well, there's quite a bit to be proud of. I mean, not just that. I mean, I'm proud of, for instance, WBAI.
Yeah, I said it. Proud of BAI because it's an American institution. Yeah, I said that, too.
And I agree with that.
I hope people call and support the station and pledge on WBAI.org or give to WBAI.org and keep the station going.
Absolutely. But this wasn't just a shameless plug for the station.
I'm fascinated. Kyle and I have been watching the feeds pretty much nonstop since last week.
The thing went up right before our show last week.
And ever since then, it's been one amazing thing after another.
But mostly, it's just the joy of the geekiness, you know, just people who are into the science and experiments and just spending a lot of time figuring things out.
And you just you hear it in their voices. You see it in their faces as they're piloting this this spacecraft as people on the ground are making sure their needs are taken care of.
It's it's fascinating to watch. It's it's one of the really good things that I think we have done in this country.
And I hope we do a lot more because, you know, science is is the future.
Science is pretty much what it's all about.
And it's just it's nice to see a time where we're not at each other's throats, where we're not trying to figure out ways to to harm people or to hurt the other side or whatever.
However, this is pure exploration and we need a lot more of it in a lot more fields.
And I don't think you can ever have a better demonstration of what we've seen.
But are there any any highlights that you can remember from this last week?
I mean, I'm looking at a story now about Artemis to astronauts making a long distance phone call to the space station.
They called the space station from the Artemis craft.
One is about 25 miles above Earth.
The other was about how far?
250,000 miles away.
So quite a quite a difference in distance.
But they were able to make that connection.
And it was just, you know, you know, we're calling back when a long distance phone call meant something.
Well, it still means something if you make it far enough.
Yeah, I mean, it's it's pretty incredible, right?
I mean, and I think that there's a lot of overlap between this sense of curiosity, this childlike attitude towards the world and the universe and how hackers approach the world as well.
I mean, I think there's just so much in common in terms of the scientists at NASA and the hacker community, not just a love of science and a love of technology, but I think a love of knowledge and and the love of sharing that knowledge as well, because that's what they're doing.
We're not we're not going into space and amassing knowledge and just hoarding it.
NASA is doing what what hackers do.
They they acquire this knowledge and they spread it around the world.
They shout it from the rooftops.
They let everybody learn from it.
And I think that there is that that's pretty cool in terms of the overlap of our of our two cultures.
In terms of the highlights, I mean, I got to say, getting the images back was pretty amazing.
Wow.
I think a lot of people with new background or or desktop images on on their machines, you know, seeing the Earth set was pretty amazing.
Seeing that eclipse was incredible.
Just being up in space and knowing that humans were farther away from Earth than they ever have been.
And we set a record and going around the moon and now they're on their way back, hopefully safely.
It's pretty, pretty, pretty goddamn cool.
And hearing them struggling with things like their USB-C connector isn't working for some reason.
It just is so familiar, except they're way out there in space dealing with these things.
But it's somehow comforting, you know, to see the four of them interacting.
And, you know, I guess I have to say, as a witness to the first time this happened, where people were sent up there, it feels like it feels like they've changed as far as who they are as human beings.
It's not just a bunch of white guys with Midwestern accents being sent up to space and coming back down and saying the same things.
You know, this is a diverse group of people.
You saw them get really emotional when they were naming a crater after the late wife of one of the astronauts.
And, you know, I never saw four people hug in space before.
Now, if this was Star Trek, all the Trekkies would be saying, oh, this is too woke.
You got to get it off the air immediately.
Not all the Trekkies, obviously, but, you know, the conservative ones.
And here, you know, this is real life.
People have changed.
People have gotten a little bit more aware of other people.
I'd like to think that.
Anyway, I know there's certainly a lot that we have to fix and mend because in the background of all this, of course, you have threats to take out entire civilizations and things like that.
And you just wonder, are these guys going to, you know, make a left turn and head to Mars instead of coming back to Earth?
But it's just, it's one of these magical moments that basically is going to last up until Friday when they splash down.
And I'm just glad that, you know, we're all around to see it and hopefully appreciate it.
If you have an interest in this kind of a thing, it's not too late.
You can just turn on NASA TV.
If you don't get NASA TV, they have several channels on YouTube without commercials.
Google does not interrupt them with commercials.
And you can simply watch what's going on inside the spacecraft or see a picture from the outside of what it looks like.
It's, you know, what they call slow TV, I guess.
Not much is going on.
And then all of a sudden, a lot is going on.
So it's really, it's an incredible thing.
Young and old, I think we can all appreciate that.
And people who are into technology and science and hacking and things like that.
But, yeah, especially.
Kyle, I know you've been interested in this as well.
Yeah, what can I say that hasn't been said?
Those are great observations.
And I'm sure a lot of people are really spellbound by it.
I think it is good.
It's a very positive development, all kinds of wonderful discovery.
And you do certainly appreciate the simple things of day-to-day life when so much of your environment changes and you're in a vehicle doing this kind of science and exploration.
So, yeah, I think for me, I have a difficult time.
I think there's some sadness, too, because you contrast it.
But I don't want to dwell.
So, yeah, sure, it's super happy and we're all the better for it.
And, yeah, I just hope that the things that other than the sort of shot in the arm and all that good feeling and sort of voyeurism we get of like watching this story unfold
and being there along for the ride sort of virtually is that the benefits of this materialize for humans on Earth faster than half a lifetime, maybe several lifetimes.
Because I think that's a lot of, you know, alongside the sheer accomplishment and record setting and all of that.
That's the other aspect of this is that humanity will benefit and I'm sure already has from a lot of the things that it takes to get to do this kind of thing.
We learn and develop things and create industry and all of that.
But I'm really, really hopeful that our collective investment in the story of it now, as perhaps you might call it a distraction, that that investment pays and yields returns for people.
Well, if all goes well in two years, actual human beings will be walking on the moon again.
That will be incredible.
The first step towards a permanent base and also first steps towards going further than the moon.
To me, that is the pinnacle.
That is what it's all about is expanding, learning, exploring.
And if you can't do that, you know, you're spinning your tires, I think.
Go ahead, Alex.
One of the things I think is really fascinating to think about from the perspective of technology and the hacker culture, too, is you think about that computation power of Artemis this week, today, right now.
Artemis right now is flying with more processing power inside of it than the entire shuttle program had in the ground the last time we went around, the last time we went towards the moon.
Yet, it's still riding on the basis of the same ancient equations, right?
It's the same physics.
It's the same equations that we're using.
It's just, you know, we're talking about thrust.
We're talking about gravity.
And there's this kind of tiny corridor between them where Artemis is going to fly.
But the technology has advanced so much, but yet, you know, there's still this ancient component to it that is a natural law of the universe that's allowing us to do this.
I think there's something incredibly cool about that.
Oh, absolutely.
It's universal.
It's something that knows no boundaries.
And in all likelihood applies everywhere.
But we're not sure about that because we haven't explored everywhere.
There might be all kinds of different rules in different parts of the universe.
This is how we expand our knowledge base.
This is how we get access to new forms of technology.
So much did come out of the space program previously, satellite technology of all sorts.
Go ahead, Alex.
You know, you've been around for a little while, Manuel.
You remember historical things, right?
Do you remember an incident in the 80s?
I think it was like the late 80s where the CCC, the Chaos Computer Club, had hacked into NASA's VMS systems.
Do you remember something like that?
Uh, I remember lots of people doing that, but sure, why not?
They probably did that as well.
You know, I got into that.
In my heyday, I got into, uh, the Statue of Limitations is long gone, but I got in.
Plus, I was also charged for it.
I was looking at plans for the space station and NASA, which was a different space station than what they built.
So, you know, there was no space station at the time I was reading this, and it was just kind of incredible.
I wish I had kept it.
It would have been pretty fascinating.
You should have.
But what about breaking into, uh, into NASA back in the 80s?
Yeah, well, it was the CCC.
It was the Chaos Computer Club that did it, and it was something called the SPAN Network.
I think it was like the Space Physics Action Network or something like that.
But it was a bunch of VMS systems strung together.
But what I thought was cool about that was that, um, you know, rather than trashing the systems or releasing information or something like that, like the CCC actually told NASA about this and alerted them to it.
And it was kind of that same spirit of sharing knowledge and helping them to further secure their systems, but apparently not secure enough.
Well, I mean, I don't think they've been compromised too much recently.
But the thing is, you know, we're all curious.
We're all, uh, probing things and, uh, learning from our mistakes and learning from progress.
And I hope this is an example of that.
Anyway, again, uh, encourage people to pay attention to, uh, um, uh, see what's going on and, uh, hopefully, hopefully be a part of exploration in the future, which is so much better than destruction.
Um, okay, moving on.
Um, first of all, some scheduling notes.
We do not have overtime tonight.
Overtime is our, um, uh, podcast version of the show that follows eight o'clock on YouTube.
Uh, but, uh, we're short staff tonight, so we're going to skip that.
Uh, that'll be back next week.
And, of course, this show will be back next week.
So, uh, we're just going to do the, uh, the hour here on WBAI.
And we're going to continue to encourage people to support the radio station as we are in our fun drive.
We're about a quarter of the way through our goal for the month, which works out because it's, uh, what, April 7th.
So, um, yeah, that's almost exactly a quarter of the way through.
So let's keep up that pace and, uh, you know, hopefully we'll make the goal.
That's the important thing.
And if we make the goal, then we, we live another day.
So we continue to operate a radio station in the middle of New York with no commercials.
Since 1960, we've been doing this.
All right.
Uh, what else has been going on in the world of, uh, technology and online antics?
Oh, boy.
Um, do I want to read this story?
It's a new controversial online game.
Um, yeah, this is, uh, this is unbelievable.
Uh, it's sweeping through classrooms across the U S it's, it's called five nights at Epstein's
as in Jeffrey Epstein.
Oh my God.
Players pretend to be sexual assault victims trapped on the convicted sex offenders, infamous
Island.
And in order to win, they have to dodge Epstein for five nights to avoid being abused by him.
Oh my God.
Who comes up with it?
The rapid popularity of the game has left many parents and school administrators deeply concerned.
Um, wow.
Uh, yeah.
Well, who wouldn't be deeply concerned by this?
Uh, it's, um, according to this story, um, from NPR, uh, it's difficult to pinpoint exactly
where the game originated, but it emerged in the last few months as the justice department
began slowly releasing more of the Epstein files, uh, to spin off of a more innocuous game
called five nights at Freddy's.
Uh, and that actually inspired another iteration of the game called five nights at ditties,
which became popular around the time of the Sean Combs trial in the last couple of years.
And now we're seeing the Epstein themed iteration of it.
Um, yeah.
Um, so, um, I don't want to go into detail as to what goes on in this game, but, um, uh,
basically kids are playing this on their school Chromebooks or their school networks trying
to bypass, bypass places where, uh, it's been blocked.
Uh, so across the top social media platforms, Instagram, TikTok, YouTube, uh, it's quite
easy to search for.
Um, but boy, uh, this, um, this is, uh, makes you think about how the kids are viewing what's
going on in the world.
And again, Alex, turning to you as a parent, uh, must give you cause for concern.
Uh, this is crazy.
I mean, uh, both my kids, actually I have one of them going through the five nights.
At Freddy's craze right now, but my, my older one did about five years ago.
So the, the age that kids get into what's called FNAF or five nights at Freddy's is probably
like around eight or nine years old.
That's when they really start playing it.
It's kind of a bizarre game.
It's creepy to begin with, to be honest with you.
It's a horror game for kids.
Is this Freddy Krueger?
No, it's not Freddy Krueger.
I don't know exactly who Freddy is, but it, um, that's a good, that's a very good question.
I think it's the name of the pizzeria.
So the, the whole concept of the game is that you're like this night shift security guard
who's trying to survive five nights in a haunted kid's pizzeria where these murderous animatronic
robots and, and mascots stalk you via certain security cameras.
It's like, it's a bizarre game.
There's all these jump scares that happen.
It's actually pretty, pretty frightening, but the kids love it and they love to be scared.
I think it's kind of awful and horrible though, to hear that, you know, five nights at Epstein's
is, is something, I mean, look, it's gotta be funny.
Maybe the older kids, that's more scary, I guess, than, than the murderous robots or whatever.
Yeah, exactly.
I mean, I mean, you could see, you know, five, five nights at Epstein's, you know, being at
Epstein Island, and then you have some kind of sequel to it where you're at his Manhattan
Townhouse, and then you have another sequel where you're down in Palm Beach and you're
trying to escape Florida or something.
I mean, there's...
I'm more concerned about the copyright issues.
I'm kidding.
I'm kidding.
That's hilarious.
But somebody out there is concerned about that.
You know, you know, you ripped us off with the Epstein version of this.
Yeah, but, but what's, what's really kind of sad about this, so you have this, this, this
fictional horror world, but now it's being wrapped in, and almost fictionalized and parodied
in this, this real world of abuse, right?
So it, it turns this like jump scare franchise about a haunted pizza into this really grim kind
of tale about what was really real sex trafficking, um, and, and an operation that, you know, was
protected by certain institutions and people around the world and has, has real victims.
And so in that sense, I think it's, it's, um, particularly irreverent to, to make light
of it in that way.
But it's a whole, it's a whole CSI syndrome, you know, where it's basically, you know, something
that happened in the news six months ago, and now you're seeing it play out in fiction.
Same thing with video games.
No.
Yeah.
I'm, I'm, I'm wondering who, who made this.
I mean, this is probably like some kind of Roblox riff on it or something like that would
be my guess.
Does it say where, where and who and how it's accessed?
I don't have that information, but, uh, curious.
Yeah.
Yeah.
We can, we can interview the person with, uh, you know, all kinds of voice masking and things
like that, but, um, yeah, that's just something to be aware of.
You're a parent that might be going on.
Um, all right.
Yeah.
Wow.
Yeah.
Well, okay.
Let's, let's, let's move towards, uh, AI if we're going to really try to scare ourselves.
Um, now Microsoft apparently is aware that you should not be trusting co-pilot.
Um, yeah, because, um, what the, um, remember we talked about last week, Alex, with the typing,
you know, you know how loud that is over the air?
Sorry about that.
I said, yeah.
So incredibly loud.
Okay.
I'm multitask here.
Yeah.
Let's, let's try and focus.
Okay.
Uh, so, um, uh, despite the last update taking place in late 2025, the, uh, document for co-pilot
for individuals recently attracted new attention from people on the net.
Uh, it includes this gem co-pilot is for entertainment purposes only.
It can make mistakes and it may not work as intended.
Don't rely on co-pilot for important device, uh, important advice.
Use co-pilot at your own risk.
Not exactly a ringing endorsement of their own product there, but one that I think is,
um, fairly accurate.
Now, uh, this, um, this story comes from the register and, um, regular readers there would
not be shocked by that warning, uh, that, um, things co-pilot gets things wrong and should
not be relied on.
Um, the company itself has long acknowledged the assistant's limitations during the London
leg of its AI tour.
For example, every demonstration of co-pilot wizardry came with a warning that the tool
could not be fully trusted and that human verification was required.
Uh, the same applies to any other AI assistant.
They can be useful, but their output still needs checking, particularly on anything, uh,
consequential, uh, like medical advice or an investment plan.
Um, basically, uh, uh, a commenter on Hacker News pointed out, Anthropic does a somewhat
similar thing.
If you visit their, uh, terms of service, uh, the one from max slash pro plans from a European
IP address, they replaced one section with this non-commercial use only.
You agree not to use our services for any commercial or business purposes, and we and our providers
have no liability to you for any loss of profit, loss of business, business interruption, or
loss of business opportunity.
The commenter added, it's funny that a plan called pro cannot be used professionally.
As for co-pilot's terms of use, they may not be new, but the attention is useful for two
reasons.
It is a reminder to read the text user, uh, users, uh, so often click through, and it
underlines that chatbots such as co-pilot are neither companions nor dependable sources
of advice.
Instead, they are error-prone tools that can be helpful one moment and confidently wrong the
next.
Some in the tech industry may market AI assistants as though they put a genius in every laptop,
but Microsoft's own warning is rather less grand.
It can make mistakes and it may not work as intended.
I'm wondering if this is the way out of the AI mess, simply acknowledging that it's not
perfect, that it's going to get things wrong a lot, and that it's entertaining and fun and
might even be helpful sometimes, but we should never, ever, ever rely on it as the final word.
I think if we just agree to do that, no matter how good it gets, um, that at least saves us from,
uh, you know, the, um, AI apocalypse that we're all dreading.
Any, any thoughts?
Yeah, sure.
I mean, I, I have a lot of thoughts about this.
This is, I think, a really interesting topic, but I, I think ultimately what, what our listeners
really need to understand about these LLMs and these AI systems in general is that they're
non-deterministic, meaning that you can get different outputs depending upon the input that
you put in, or you put, you get different outputs for the same input as well.
So, I mean, that, that becomes a problem, right?
So, if you, if you think about co-pilot, right, under the hood, co-pilot is kind of like
clippy, but with a law degree, with a physics degree, et cetera, right?
It's like kind of clippy on these quantum dice.
So, if you ask it the same question twice, and you get three different answers, how are
you supposed to build some kind of policy off of that?
How are you supposed to build anything off of that?
How is that a solid foundation for anything?
But, so if you understand that AI systems are going to provide different answers at any
particular time, understanding that they are non-deterministic is the technical term for
it, that's really, really important because, you know, non-deterministic output, when it's
plugged into deterministic compliance rules or, or systems, is how you have to end up explaining
to somebody like a regulator or your, your boss or having a conversation with a lawyer or
your general counsel that, oh, you know, the model did it, you know, while you're sitting
in some kind of HR hot water.
So, I think we're right not to trust these systems with everything.
I think their rules and their, their warnings are, are accurate, but I think it's also got
to be reflected in their marketing materials, not just to find them.
Do you think Microsoft is doing the right thing here by telling people not to trust CoPilot?
Yeah, I think they are.
I think they're, I think everybody should not trust their AI systems.
And, and the one thing that AI systems do not have that human beings do have, and which
I think will never be replaced entirely because it's empirically based and experience based
and that's judgment.
You know, they, they don't exercise great judgment.
They may be able to put together an argument.
They may be able to put together a summary.
They may be able to be able to synthesize information much faster than we can in much
the same way that when, when computers first became, you know, really accessible to the
public, we were able to do so many more things with them than we could without them, namely
like count things really fast.
We put things in memory, all these things, you know, that we're going to change the world
and change the universe.
I think that's how we have to think of AI systems now as well, is that they are these helpers.
They help us automate and do these mundane tasks, but they should never be used to substitute
for human judgment.
I mean, I, I use AI systems all the time.
I use Claude code a lot, but I always have to check it.
You always have to verify the sources.
It gets things wrong sometimes.
It, um, and it exercises particularly poor judgment on, on, on hard cases where there's
gray areas where one thing could go one way and one could go the other way.
And you can iterate and you can go back and forth with it and say, well, have you thought
about this?
What about, what if we did it this way?
And then you get some other non-deterministic answer.
So you really can't trust the output no matter what.
You always have to trust your own judgment.
And when it goes down a wrong path, it's very hard to get back from that wrong path.
If it says something as a fact and you say, well, no, that's not true.
Uh, it'll keep going back to that fact and basically, uh, almost doubling down.
The thing to remember with AI, it doesn't care.
It's incapable of caring.
Whereas a human being cares, at least, you know, you'd like a human being to care.
Uh, but the other thing to remember is that humans certainly aren't infallible either.
Uh, you say, if you ask, uh, AI a question three times or three instances of AI, uh, the
same question, uh, you'll get three different answers.
Same thing with humans.
You'll get three different answers.
Humans will be wrong as well.
We seem to think that because AI is fast and because it, it can say a lot that it's always
right and it just can be wrong a lot faster.
And that's something I think we need to remember.
Here's another story that I thought was, uh, kind of interesting.
Uh, this is, uh, involving Target.
Um, they are saying that if, if their AI shopping agent makes an expensive mistake, you're the
one that's going to be on the hook for it.
Uh, now, uh, big box chains have proudly embraced artificial intelligence.
Don't expect them to help when their AI assistants charge your card for hallucinated garbage.
Yes.
That, um, that trend in which retail giants force their AI initiatives on consumers while
distancing themselves from its failures is growing.
And while that attitude certainly isn't unique to big box stores, it's increasingly becoming
their default setting, revealing their, uh, the ridiculous double standard that's baked
into the AI boom.
Uh, this, uh, article coming from futurism.com target is the latest to jump on the bandwagon.
Uh, according to business insider, the Minnesota based retail operation, uh, or corporation, uh,
recently updated its terms and conditions to make it clear that customers, not the company
would be responsible for the failures and hallucinations of its upcoming AI, um, shopping assistant.
Uh, that virtual buddy, which runs on Google's Gemini is supposed to help online shoppers finish
target runs on their users behalf under the new terms.
If a customer uses the Gemini agent to do their shopping for them, any transaction performed
by the AI would be considered transactions authorized by you, which, uh, translates to any
mistake by the AI agent, whether it buys the wrong item entirely or a super expensive version
of the right one without your consent, that will come out of your pocket.
Not without your consent, that bugs me because, you know, they're saying here that, um, they
consider transactions to be authorized by you, but if you're not giving your consent, how
can you make that assumption?
Well, the new term state, you are responsible for reviewing activity performed by your agentic
commerce agent.
Yes, that's what it's called.
And for promptly notifying the agentic commerce agent and target of any activity you believe
is unauthorized or outside the scope of permissions you approve.
So you have to be on top of this, watching every move.
Um, if that weren't enough proof, the target doesn't stand by its product.
It gets more explicit.
Target does not purport to guarantee that an agentic commerce agent will act exactly as
you intend.
In all circumstances, you should review orders, account activity, and settings regularly.
It doesn't sound unreasonable.
Uh, yeah, you should pay attention.
You should see what is being done on your behalf, but you know, it could get kind of, uh, difficult
if it's, um, uh, a slight deviation from something that you want, uh, or it's giving you the wrong
product and you were supposed to catch that, but you know, it was a mistake that the AI made
and not you.
So you had no idea it was going to make that kind of mistake.
Like I could see times where you would feel like you're being victimized.
Any, any thoughts on this?
It feels like they're offloading a lot of the liability onto the consumers.
Yeah.
That's all it is.
I mean, that's my takeaway is that they gain some sort of, um, profit or a reduction in
their overhead, whatever their business runs better by using this, by implementing this
stuff and, uh, conditioning us to use it in our consumption of their products and any sort
of problem, this is going to pave or smooth the way for, uh, absconding themselves of any kind of
responsibility for what is essentially something that is going to make the processes of their
business easier for them to do without, I guess, maybe hiring people to answer phones or answer
questions or, um, chat live, uh, while someone's trying to pick something out, whatever it is,
it just, I'm so skeptical that this is, um, uh, uh, worth the risk to accept.
I mean, I would, I would think very carefully about the kinds of things that they're, um,
that they're kind of, uh, putting in your, uh, putting in your responsibility, like putting
offloading onto the consumer.
Well, it's interesting that you say, you know, uh, hiring people to help out.
Imagine that, imagine if they had to do that, these, these big companies that used to hire
people to help you replace them with AI.
Now they're saying AI is not people, you know, it's, it's, it's not, uh, something that you
can rely on.
Uh, well, okay.
Bring back the people then.
I think that sounds like a good idea.
It's basically, you know, they're trying to cover their asses.
There's no question about that.
Well, and I think they think, they think that we'll accept it because it's this sort of temporary
thing because off in the future, we're going to have this general intelligence that doesn't
make any of these mistakes and they'll just upgrade whatever they're bolting on to their,
um, commerce sites now to that.
And, um, we won't have these problems then.
And I don't know.
It just sounds like geoengineering out of man-made problems, you know, or I don't know.
I just, it, it, it, I, I can't buy that this is all for the benefit of the consumer and that
we, this is why we should be thankful for these, uh, brand overlords for, for implementing
this, this fancy technology.
Um, I got news for you people, uh, high tech new technologies.
They're not going to be brought to you necessarily, uh, as a gift from these companies.
They'll, yeah, sure.
They'll sell you their version of it, but, um, ideas, innovation, um, uh, it is probably
not something that is commodified.
It's something that, um, you might be better suited at doing than a lot of these, um, firms.
It just, it's, it boggles my mind that like people can adopt free services and do these
sorts of, uh, embrace these things that are essentially magic to them and, um, think of
it as like the pinnacle, like the, the end all of, of, um, high technology.
I, I, I think that's really selling this era short.
Um, there's so much more you can do than, than what pops out of a vending machine.
Well, Target is joining other retail giants like Walmart, uh, in rolling out AI shopping
agents that it openly distrusts, which is pretty fascinating when you think about that.
Uh, Walmart, which recently, uh, recently outraged consumers with AI powered price gouging
also updated its terms of use to cover for mistakes made by its AI shopping assistant.
Who can name Walmart's AI shopping assistant?
Anyone?
No, no, no.
Kyle, I don't know.
Uh, well, I do.
I have it in front of me, but I didn't know before now.
Wait, is it Wally?
No.
Oh, that, that would have been a good name.
I don't know why they didn't use that.
No, it's Sparky.
It's Sparky.
Yeah.
Sparky.
Do you remember Sparky was the name of the, uh, Spark station at Stony Brook that everybody
used to hack back in the day?
Well, I don't think Walmart knew that.
No.
Kyle.
Well, I will say that, um, this is an area of interest.
I mean, the ability for these things to find different products, to go through what's available
at different e-commerce sites.
I've played around with it.
I think, um, a lot of people are realizing that this, this could be very, very helpful,
but not everyone is the same and not, you know, and not all of them are, are equal.
And you have to understand that, um, if you're using the, the implementation of it from a
particular entity, it may be getting very little in the way of resources and maybe heavily
weighted towards, um, recommending and pushing certain products and certain things on you
that you might not be aware of other deals or competitors or other ways of doing something.
And it'll be biased in, in favor of the firm that's, that's implementing it.
Uh, but if you use, um, what I think of as search engines, but some of these assistants
to, uh, speak to them naturally and ask them questions about products and things, um, to sort
of augment your insight into the marketplace and your ability to shop, I think there can be
huge benefits to that.
And, um, it can be fun to play with.
There was an instance, I was, I was looking, uh, um, I was installing something that had
dip switches, like a, an electronic device, a controller, um, for charging, um, and battery
set up.
So anyway, um, it was telling me about all of them and it was asking me a little bit about
the, the model that I got, cause I was upgrading it to a higher, um, uh, power rating, but it,
it, in the process, it was so confidently telling me that, um, yes, these have to be off and,
you know, all eight of these dip switches have to be off.
And then in parentheses, it put the position of the dip switch and it said up, but up was
on.
I was looking at the manual, like the paper manual that I had, and I'm looking at a diagram
and clearly up is on.
So if I had followed, so it was contradicting itself in the same list of instructions.
I didn't even bother trying to correct it, but that's the kind of stuff where you have
to know a little bit more about the scenario.
You can, it's fine.
You know, you can, you know, load it with, um, schematic or information like facts and
help you go through things.
But it's not, um, unless you're in control of it, parsing things, you know, step by step
and stuff, you're really leaving it to make all kinds of these mistakes unattended.
So you shouldn't be surprised.
You have to take the high road sometimes.
Absolutely.
It's going, like I said before, it's going to get things wrong and it's going to double
down and keep getting it wrong.
And you can try and correct it as many times as you want, but it's still going to, uh,
keep assuming something that's not actually true that you know, isn't true, but there's
just no way to convince AI.
I didn't even, that's exactly why I didn't even engage because I had heard, uh, accounts
of you trying to correct, um, something, you know, that you noticed that was a mistake
to see if it would, uh, atone and correct its own answer instead of defending it.
I didn't want to get into that.
I just knew, okay, this is wrong.
I know I got what I needed out of this as far as steps, things to look out for other
tips that, that passed what I felt like was a logical, reasonable things, little reminders
that I might not have been thinking about, but take it for a grain of salt.
Absolutely.
Have, have you had a situation where you're online trying to fix a problem, uh, with something
that, that you, um, uh, subscribe to or have bought and you're talking to the company and
you know, you're talking to AI and you know, you're making some progress and maybe it even
solves your problem, but sometimes it doesn't.
And then you wind up talking to a person.
You asked to be transferred to a person and now you're talking to, to, uh, what was her
name again?
You were talking to somebody, it was the exact same name, Kayla or, or Shana or, oh, uh,
Kyla.
Kyla.
That's right.
Don't knock Kyla.
Kyla was very helpful.
I've talked to Kyle with a totally different company.
And I just wonder why do we keep getting the same names?
But yeah, so we're talking to this alleged human and, uh, now we're talking to somebody
who is, is able to solve our problem on a human level.
But how do we know?
How do we know when I'm talking to another AI, how do we know the first AI wasn't just
a fast typing human?
There's no way to know.
There's no way to really know who or what you're talking to.
I just, it's, it's, it's, it's both maddening and hilarious at the same time.
Alex, go ahead.
Well, I, I think that's where this whole notion of proof of humanity comes into play.
And, and we need a naming system for these agentic helpers as well, so that they can
be perhaps discovered or contacted or something like that, that there's lots of standards and
various things being thought of at the, at the moment with respect to the discovery and
interaction of these agents amongst each other.
But, but to go back to this issue of, of trusting them and in particular with Target and Walmart
and all these, these other companies that are, are distrustful of their own agents, helping
people shopping.
I mean, I think it goes back to the, what we were talking about before that it seems like
Target and Walmart and all these other big retailers actually do understand what's happening
under the hood with respect to these AI systems and the fact that they are non-deterministic
and that it might recommend you to buy a Rolex or it might recommend that you, or it might
recommend that you buy a Vacheron or, or, or something like that, depending upon how you
ask it or the particular time of day on which you ask it.
But the non-deterministic nature of this, I think is really interesting.
It seems like they understand it.
But to my mind, I think there's something really exquisitely perfidious about a company
in an AI system that begs you to trust it and use it.
And then immediately hides behind some kind of legal disclaimer when you're unhappy with
it.
It's like, okay, come on.
You have to stand behind your product.
Either, you know, it's fit for purpose or don't offer it at all.
Well, speaking of, that's what, that's what.
Speaking of standing behind your own product.
May 20th, Amazon will stop supporting any of its Kindle devices that were released in
2012 or earlier.
That means that these devices, even if they're perfectly functional, won't be able to connect
to the Kindle store to download new eBooks.
Amazon says affected devices include the first and second generation Kindles, the Kindle DX
and DX Graphite, the Kindle Keyboard, the Kindle 4, the Kindle Touch, the Kindle 5, and
the first generation Kindle Paperwhite.
According to The Verge, Kindle Fire devices will also be affected on the eBook front, though
other apps will still work on the device.
Yeah, so this is outrageous.
Yet another example of how these companies just expect you to buy it again.
Other recent examples of companies doing this.
Spotify ending support for its car thing in 2024.
Google discontinuing its older Nest thermostats last year.
That affected me.
In February, Netflix ended support for nearly 87 million older devices.
I've had that happen in the past.
You know, they still work.
They still do everything they're supposed to do.
The company that sold them to you, they still exist.
They're still selling the service.
They just don't want to sell it to you anymore.
They don't want you to benefit from it.
But you've already paid for it.
It's not like you're paying a monthly subscription and they're saying, oh, the subscription fee
is going up or we're discontinuing that subscription.
No, you've bought the device.
And the understanding of buying the device means you will have access to this.
If they're going to change the rules, they need to provide you with a new device to replace
the one that they're deciding is no longer worth it.
I mean, what if you went down to Amazon and you tore down their building and said, this building
is old.
It's no good anymore.
You need to build a new one.
They wouldn't like that, would they?
Now, you know, consumers don't like to be robbed.
And this is robbery.
This is absolute robbery.
Yeah, you can laugh at somebody who has a Kindle that's 14 years old.
But, you know, I actually think somebody who has a 14 year old Kindle is worthy of respect,
just like people who have old phones, old computers, old bits of technology.
They've kept them running and they know what they're doing.
And they should be allowed to continue to run their particular hardware or software.
These companies should not be interfering and saying, you know, you can do it better and
we're going to force you to buy it again.
That is just completely wrong.
And I hope there's some kind of reaction against this.
Even though I don't own a Kindle that old, it's getting close.
And this kind of thing just really annoys me.
Kyle.
I just would point out, I agree, Ben.
My attitude is that it ought to be illegal to have these kinds of business practices with
products that are themselves not defective, but rather becoming an end of life as far as
the support from the manufacturer.
I think it should be illegal unless you provide some form of offer, like what you're talking
about, rebate or some sort of buyback or replacement options that are really competitive and don't
make the consumer feel like they've lost value or lost capability from what they originally
purchased or some other combination.
I think there should be recourse for sourcing parts in the aftermarket when the manufacturer
stops supporting them or up to including schematics or specifications for things so that if people
wanted to have a large purchase independently for perhaps battery packs for things that need
replacements, et cetera, like that stuff is basically the heart of a lot of right to repair legislation
that is catching on in this country.
And, uh, I just, I think that, yeah, if you like your device, um, you don't want to, uh, make it a source
of waste immediately, you should have the ability to repair it.
You should have, uh, options for, for, uh, at least some phased option, um, to take ownership of it when
there's no one else to help you with that support.
I mean, look how they have us trained.
You know, we all have phones in our pockets now, but that's not good enough.
You know, you pay hundreds of dollars, sometimes over a thousand dollars for these phones.
You'd think that would be enough.
You know, when I grew up there, we had a phone, you know, the phone was there.
It was dependable.
It rang, you picked it up, you talked.
Yeah, it was simple, but it worked.
And, you know, it worked forever.
Here we have something so much more advanced.
You would expect it to last a while, but no, five years max.
And then your phone is considered obsolete.
You have to buy it again.
You have to buy another one or, or trade it in and, uh, you know, pay a few hundred dollars
more.
But you're in this cycle where you just have to keep buying and buying and buying and paying
the same companies.
And, you know, at some point, why can't you be happy with what you have?
And, and, you know, be satisfied that this phone is going to work this year, next year
for the next 10 years.
But no, we can't even wrap our heads around that.
A phone that lasts 10 years.
My God, what, what kind of dinosaur are you?
Well, you know, maybe, uh, the, the, the newest app isn't always what you need.
Maybe you want the basics.
Maybe you want to just be able to make phone calls or text or, or, uh, play certain video
games or whatever.
It's up to you, not up to them.
They have us trained so well that it's abnormal to think otherwise that you have to buy a new
phone every five years.
It's, it's really ridiculous.
Alex.
You know, what really gets me about this is if you think about the, the dirtiest part of
the tech life cycle, it's not, it's not necessarily the power that you're drawing from the grid to
run these devices.
Think about it in, in these terms, right?
The environmental impact here comes from the perfectly good hardware that these companies
are pushing off a cliff and the name of the support life cycle.
And then these very same companies then dare to run some kind of audacious earth day campaign
on the way.
I mean, that's, what's really unbelievable to me.
This is where I think we need to call them out is like, let's wait for earth day to bring
up the right to repair next time.
I think those two are, are perfect compliments of each other.
Yeah.
And I mean, we also should just educate ourselves.
You should have work parties.
We should have a repair parties.
Like all of these kinds of things should be something that we culture ourselves to step
up to the level of engineering that is all around us and not be so disposable.
Yeah, absolutely.
We're out of time.
Unfortunately.
Oh, I was going to talk about the, they, they found out who started Bitcoin.
Oh, can we wait till next week?
Oh, you don't want to talk about crypto?
Let's do this next, you know, we'll remember next week.
Don't worry about it.
OTH at 2600.com is our email address.
Please write to us.
Support WBAI.
Give to WBAI.org.
Become a BAI buddy.
Keep 99.5 FM alive in New York City.
This is Emmanuel for Off the Hook.
We will see you next week.
Keep looking to the skies and keep watching the astronauts.
Good night.
The hatch is coming open.
Okay.
It's open.
Terrified.
I'm moving outside.
Okay, I'm letting go.
Copy that.
I'm starting to drift away.
It really is dark out here.
How am I doing?
You're doing fine.
I'm beginning to see a point of light on the horizon.
It's getting alert.
I'm getting some pretty good oscillations.
Oh, it's coming quicker now.
It's really going.
Any second now.
Here we go.
And, uh...
Oh, I mean, I mean, I...
That there really is something.
Oh, it's tremendous.
Tremendous.
We'll be right back.