PHP2Shell And The AI That Found It

Jason Howell
00:00
Coming up, people are using Claude to make their smart TVs dumb again. Also, synth nerds. like me, although I'm not vibe coding, but a lot of them are vibecoding new firmware to flash onto their old gear.

00:15
Really interesting trends going on right there Researcher who helped name superposition now says we were looking at how neural networks work underneath the hood all wrong and Disney wanted character AI to go bankrupt before it would sign the check. Is that okay? Well Jeff Jarvis and I talk about it next on this episode of the AI Inside Podcast.

00:49
What's going on everybody? Welcome to AI Inside, the show where we take a look at the AI that is layered throughout the world of technology. I am one of your hosts, Jason Howell, joined as always by Jeff Jarvis on the other side of the United States of America

Jeff Jarvis
01:03
Hello, hello. And I was just saying hello to Daniel Croft, our wonderful uh friend and fan from the other side of the ocean, I think, right, Daniel?

Jason Howell
01:09
Yes, yes, indeed. Yes indeed. Daniel, thank you so much.

01:13
He helps us behind the scenes with shorts. He's always in here in the live show making us feel like we have at least a a small live audience. We actually we do get a decent amount of people watching live between LinkedIn and uh the YouTube channel and the other places that we stream.

01:29
We just don't get a whole lot of chat action during, but um but hey, we got Daniel in there. We got Jesse in there, it looks like. Hey Jesse, welcome.

01:37
Good afternoon. Daniel's in London. So, you know, we'll take what we can get.

01:42
It's good to see you folks.

Jeff Jarvis
01:43
A friend of mine just uh sent me a picture from one of my favorite bookstores on earth, the Blackwells in Oxford. And there was a stack of hot type there. It made me very, very happy to do that.

01:53
Yay. And by the way, since I've never blogged go ahead and go to JeffJarvers. com where, folks, there are still signed copies of the Mockler Brooksletter.

01:59
So you can click on that link and get them.

Jason Howell
02:01
There you go. Jeffjarvis. com, the newly redesigned for Jeff Jarvis.

Jeff Jarvis
02:06
com. Was surprised to see last week. My son Jake did it on uh On the sly. Did a great job.

Jason Howell
02:13
I love it. I love it. So you can get access to all of Jeff's amazing books there. Maybe Jeff will write a book About you can see everything. The new documentary.

Jeff Jarvis
02:24
There's been a lot written about that.

Jason Howell
02:26
There has been at this point. I bet people are getting pretty sick of hearing about it too. We don't have to talk about it at length, but I will say Fascinating film.

02:36
Never seen one like it. Like it is truly unique and incredibly compelling. And I my my urging to anyone who hasn't seen it, who's on the fence or whatever, is It's probably going you you're probably going to get a lot out of it, even if you don't care about the Theranos story.

02:53
It's it's kind of only tangentially around Theranos. It's more about the human condition and and what what defines truth And um it is such a cinematic experience. I I recommend you go see it now, not later, because I think it really

Jeff Jarvis
03:09
stuff's gonna leak out. Yep.

Jason Howell
03:11
Stuff's gonna leak. It's gonna get continue to be hyped to the point to where it might be one of those films that like by the time you watch it later you've heard so much hype that it couldn't possibly live up to it. But I thought it was top three film of the year for me. It was such a great experience. I don't know how you felt about it, Jeff.

Jeff Jarvis
03:28
I I loved it. Um by the way, I'm putting in another link here before I forget. So um I think it's an amazing portrait of deception.

03:40
And it's not in, I was thinking about it this morning, thinking about talking about with you now. It's not as if somebody decided to stand back and do a do a haughty fig piece on deception. Instead, it just puts you and it's not it's yes, of course everything is staged.

03:54
That's the point of deception. But you don't really know what is. And so it's about her deception, obviously, with her shareholders, her employees, her um the patients who are involved.

04:07
uh uh and and so on. It was about um

Jason Howell
04:10
and herself.

Jeff Jarvis
04:12
Yes, and herself and her relationships. And then there's and then there's Nathan Fielder who just does a brilliant job on camera.

Jason Howell
04:19
He's great.

Jeff Jarvis
04:20
He's great. And um uh you know, he's he's then making something about the deception of the form of the documentary.

Jason Howell
04:28
Yeah.

Jeff Jarvis
04:29
Uh we we call it a documentary, we say it's facts, well what do we know? Is it really? Uh

Jason Howell
04:33
that's and that's kind of the point is is like yes it's a documentary. But but that's that's what I mean by a film unlike any that I've seen, or probably unlike any documentary that I've seen. Because it really is hard to to pull out like what is documentary here and what is truth and what is you know it It's really fascinating.

04:53
I know there are some folks in the in the chat room I didn't uh that haven't seen it. They're saying no spoilers. We won't spoil you the because this is a film that you do not want spoiled.

05:02
No. Um But I didn't realize that it's not out in the UK until mid-November. Why is that?

Jeff Jarvis
05:07
That's torture. That's just that's just ridiculous.

Jason Howell
05:09
That's super torture.

Jeff Jarvis
05:10
Haven't they learned that we're one big world now? Um it's it's awful. Uh so I I appreciate that. And then the other countries where they're waiting for translations or whatever, it's gonna take forever. Um right, and then there's other deceptions that come or other questions about deception that come which I can't tell you about right now.

Jason Howell
05:26
Yeah.

Jeff Jarvis
05:27
FYI, uh Paris Martinot and I uh decided we were gonna make a little kind of mini podcast, Ebert and Siskel, oh about this. And then and then it got into a bigger thing, so it's on the club at Twit. Uh and uh Mike is joining in. Uh anyway, we're doing it tomorrow evening. We will put up with the spoilery

Jason Howell
05:46
is definitely for people who have seen it

Jeff Jarvis
05:47
Because we want to be able to discuss it in full. We

Jason Howell
05:50
just want to discuss it. That's the thing. Yes, totally. So we'll we'll discuss

Jeff Jarvis
05:53
there's a few things. I'll I'll I mentioned one scene very early on where I said this is different. Uh and it's a small scene and other many other people have caught on this.

06:01
Uh in in their in their home um uh on the Pacific Ocean, uh ratty little beach house where the whole thing is done. Pardon me. Uh Elizabeth is uh mixing eggs.

Jason Howell
06:15
Yeah.

Jeff Jarvis
06:16
And uh Billy, her life partner, goes over and takes the bowl from her and starts mixing them. And that was a telling little moment. But what was amazing to me was that Nathan Fieldler who s who was there, you know, watching all this as part of this, he said, why don't you do that?

Jason Howell
06:31
Yeah. He calls him out.

Jeff Jarvis
06:33
He called it out. And I wouldn't have done that, right? I would have been embarrassed. I would have just watched it. I just would have, you know, thinking I was a journalist, I would have kind of been from the distance, well I noticed something and I can write about that.

Jason Howell
06:41
Of course.

Jeff Jarvis
06:42
But he just came out and said, why'd you do that? And um it was a very illuminating moment and the movie is filled with those kinds of personal moments. Um uh if you're gonna take your kids you should be warned that you're gonna see more of Billy than you wanna see in a very quick shower scene. Um uh it's it it's it's an amazing puzzle, a big puzzle that was put together in a fascinating way.

Jason Howell
07:06
It really is. And obviously,

Jeff Jarvis
07:08
why are we talking about it here? It's about a tech company. And it's about the BS of Silicon Valley and about venture and about how people convince themselves of things.

07:17
They they dis as you said earlier, Jason, they deceive themselves. She deceives herself. all the VCs deceived themselves, the board members deceived themselves uh in this whole process.

07:27
And that's yeah. I got that. I also finally watched uh the dropout.

07:32
Which is the miniseries about her.

Jason Howell
07:34
Oh, and I haven't seen that yet. Now now I'm compelled to see it.

Jeff Jarvis
07:44
George Schultz's grandson, he was a whistleblower. It split up the family. That's all in the dropout.

07:49
That's not here. But Tyler Schultz is out there doing interviews about this because he was a whistleblower. And so having the further background I I found useful.

07:59
Uh Paris said she went to see it twice because knowing what she knew at the end, she wanted to Judge of the beginning and see it.

Jason Howell
08:06
Oh, I will I would love to see it again. My wife did not see it. Um I it was a friend of mine who is uh is a tech uh tech writer that lives here in Petaluma and he reached out Saturday morning.

08:17
He was like, no one in my orbit uh knows about or cares about this except possibly you, would you like to go see it? And I was like , yes, sign me up. He was like you're the only tech guy I know in this town that would care about this movie And uh we went and I got to the end of it and I was like I gotta take my wife to this.

08:37
So did

Jeff Jarvis
08:37
did Stacey just say I know I'm not interested or I'm busy?

Jason Howell
08:40
She yeah, she knew very little about it. I mean she said what uh actually I've heard a few people say, which is like I really don't care about the Therano story. And I'm like, you don't need to Like that's not what this is about.

08:50
This is phenomenal. Yes, it's about it, but it's yes, it's way wider than that. fascinating and it gives me goosebumps thinking about it because there aren't many times where I go to a movie and I see something that is authentically different where I go Who whoever made this, I am now like full convert, can't wait to see the rest of this person's films.

09:18
And I feel that way about Nathan. He's got such a unique perspective on this stuff. And Some of the stuff that I've seen ahead of this is incredibly unique.

09:26
I think what Nathan did really well, and then we'll move on, because I really don't want to spoil anything for anyone, um, is he did a really good job of being in the position of you, the viewer He would ask the question that you, the viewer, ask when you see that moment on this on the screen. Yes. See that thing happen and you go What the head was that?

09:46
And then he's there going, what the hell was that? You know what I mean?

Jeff Jarvis
09:49
But the difference I think again is that and I'm a journalist, I'm supposed to ask difficult questions, right? I wouldn't have had the guts to ask many of the questions he asked.

Jason Howell
09:56
No, totally. That's that's what makes him so compelling to me because he he has no filter. Like he is he has no wall there. He is okay going there. And I would I would have a hard time with that too. So

Jeff Jarvis
10:09
let's go let's also mention very importantly the director, Lance Oppenheim, um, who's brilliant, uh, who who uh did um Sperm World, The Python Hunt, Some Kind of Heaven. Oh, that was he was editor there and other things. And uh Paris is a huge fan of his.

10:26
So she's seen more of his ow. So I think that's important as well. And one more thing to mention, the way we'll move on.

10:31
Uh I put up a link, uh, the third line there, is a video that Hank Green made. Hank Green, the great explainer of all things, about how blood testing works in in the in in the inevitable fashion that only Hank can do it. And it's brilliantly done.

10:46
It's an it's like an hour and twenty minutes. It's a long video But then he uh interviews Tyler um uh Schultz, whom I mentioned a few minutes ago uh in a long and really revealing uh interview. Now this was recorded before in anticipation of the movie but before um Hank had seen it So there's no spoilers in it, because Hank didn't know anything.

11:11
But it's really educational and it really helped me understand why what she was claiming was such BS. That we can take just as much blood and we can take it as a finger prick. And you know, he goes on, for example, he explains why fingerprint blood is worse as a sample for all kinds of reasons than blood out of your vein.

Jason Howell
11:33
Oh, okay. Why

Jeff Jarvis
11:34
then no, the amount isn't a big deal. He said that we we we still take the vials of blood because we've always taken that much. We don't they don't really need that much, so it's not a question of amount.

11:43
And then he explains how the tests work and how they get and it's and it's just magical, wonderful science that Hank makes clear. So after you've seen the movie Uh um before too. Maybe maybe it actually helps before.

11:55
I I definitely recommend watching this. This is just this is really great. Hank, I'm a huge fan of Hank and By the way, he did made a very nice video at which the Gutenberg parenthesis was the was the punchline of the whole thing.

12:06
And so I'm a little more grateful to Hank Uh he's just the greatest. Both of them are the greatest. Both brothers.

12:11
So anyway.

Jason Howell
12:12
Cool. Now on

Jeff Jarvis
12:13
to the present tense.

Jason Howell
12:15
Yeah, I'm gonna pause this video of Hank Green and put it over on my other screen and keep it up so that I remember to watch it. Because sometimes we talk about these things, I'm like, I'm gonna do that. And then I close the window and I forget about it forever.

12:29
So I w I do want to watch that. Um so that is uh hopefully spoiler free enough. I think we did a good job avoiding spoilers.

12:38
Uh except you now know what the front the full frontal nudity actually was. But anyways

Jeff Jarvis
12:43
, that was just a week.

Jason Howell
12:46
PHP getting into AI uh news or related to AI anyways, PHP uh news dropping pretty hardcore security um advisories happening um this morning. I don't know. Did this start happening last night? I and I I didn't pay attention. Oh

Jeff Jarvis
13:03
let me see here.

Jason Howell
13:04
It feels very late breaking

Jeff Jarvis
13:06
um the PHP Foundation. put out uh something on the twenty first explaining exactly what happened and and they they they knew that this they knew that the AI had found the faults and they were working on it and then there was a leak of that and uh so they had to kind of rush out the fix. They have put out fixes to BHP and they're telling everybody install these

Jason Howell
13:29
Yes. Okay. So PHP, um which you know, this explainer is f for as much me as it is anyone else who's like, okay, I know I've heard that before, but what the heck is it?

13:41
It's basically the language running on the server behind a large portion of the web. So you got WordPress, Drupal, uh Wikipedia's software, Joomla, um, you know, supposedly powering up to or around 70% of websites. So a lot of websites, one researcher put it at 1.

14:02
3 million sites, vulnerable to this thing that was discovered, which basically PHP is pushing out an emergency set of security fixes. uh to address two critical vulnerabilities. Uh PHP2 shell is what the pair is being called.

14:17
And this is unauthenticated remote code execution. So anyone can hit a server with no login and potentially run their own code on it um affects all PHP installations serving web requests. So a a pretty significant security vulnerability to say the least on a wide swath of sites that doesn't really require a whole lot um to do it it sounds like Um and open AI, why is this on an AI show?

14:50
Because they say that the security firm Caliph found the flaws using GPT-5. 6 SALL. Wiz, another firm, found the same bugs independently.

14:59
They say they use Gemini 4 Argon and GPT 5. 6 Sol. And um, you know, this is just another one of those examples where uh you know something that would have taken months of work by you know massive uh world class teams seems to have been uh carried out and recognized uh in a matter of uh in a very small amount of time because of AI models and their analysis capabilities.

Jeff Jarvis
15:28
Which in the short term is disturbing because oh my God, there's more weaknesses, but in the long term, I think we're going to end up with a hell of a lot better bulletproof software. At least to this extent. So the PHP Foundation put out a notice explaining all this in addition to the fixes, and they said that on Tuesday, October 20, around three o'clock UTC There were two um uh vulnerability uh uh IDs circulating on online forums.

15:53
It came with a strong with a wrong release time and a wrong score. In response, the PHP project brought its pre-announcement forward and published it at 7. 30 UTC.

16:04
It also moved the release of the fixes from October 22nd. to October 21st at 2 p. m. UTC. I guess I'm not supposed to use PM and AM UTC, but whatever, folks.

16:16
Don't don't go patent all pedenter on me. Um distributors were still packaging the fix, so releasing immediately would have left many users without an easy way to update. So moving the release forward by one day was the best compromise I'm reading here.

16:28
between acting quickly, giving people time to prepare, and making sure everyone had a simple update path when the vulnerabilities were disclosed. Okay, but then as a result of the leak and pre-announcement around eight o'clock uh UTC one of the security firms that had reported the issues mistakenly published a video. It showed their proof of concept working against a number of targets and revealed a few additional, though limited, details of the vulnerabilities Copies of the deleted video started circulating on social media.

16:58
The video contributed to an accelerated development of the various fixes. And then that's why they put it out today, the 21st. So everybody's trying to be absolutely the best they can be. Things happen.

17:12
There's a risk. But if if if you're a developer, you know what to do. The host, you know what to do.

Jason Howell
17:20
Youch. Yeah. UK, Canada, Australia, CERT, EU urging sites to go offline until you can update. Um, you know, a number of sites, including Finland's government and the Dutch House of Representatives, pulled their sites offline, apparently. Um yeah, so some exploit attempts s happening um before ahead of the patch and

Jeff Jarvis
17:44
or or the fear that they would happen. I don't know that they 're not.

Jason Howell
17:46
Yeah, right, right, right.

Jeff Jarvis
17:48
So all in all again, I think this is good news. Yeah, I discovered a vulnerability that could have been discovered by a bad guy, could have been exploited by a bad guy at length, was found, was brought to the right people, was fixed, but then there were just a few minor screw-ups along the way

Jason Howell
18:06
Well, and so now, you know, again, we're we're kind of left in this uh this position of the hmm. But just the fact that these these models are so good at recognizing this stuff or or seemingly uh much better at recognizing these stuff at a at a faster rate than humans and the the problem like On one hand, you want to know about this stuff, so you're so you're happy that you have the ability to do that. Um on the other hand, it Bio has the potential to create a large downpour of of holes and and uh openings, at least in the near term potential, you know, ways of of exploiting um the the rate and the speed at at which things are are recognized right now.

18:52
You know, being caught flat footed is a is a thing and this is uh uh has wide reaching uh impacts. And so, you know, do people have what they need to address this in a timely fashion? when these models are so good at recognizing these things.

Jeff Jarvis
19:10
I mean yeah, I mean the next release of PHP, I'm sure they're gonna put it through the ringer uh on these models and even before the release they're gonna find it, but even after the release the right thing would have happened here is, oh, we have a new uh version. You have to put you have to install now. We're telling you to install this now.

19:26
We're not telling you really why. And you won't know why. And we're not going to share the stuff with the vulnerabilities with the bad guys until everybody's installed it.

19:35
And that's the way it should happen. It just didn't it was out of order here.

Jason Howell
19:38
Yeah, right, right, right. So, okay. There you go. It seems like a lot of a lot of activity happening today around this, and it was a little bit of a late addition. I didn't have a huge amount of time to pour too deeply into it.

Jeff Jarvis
19:52
And it's complicated, too.

Jason Howell
19:55
Yeah, it's just i again, it's like PHP is one of those things like I've been to in technology a long time, but there are these terms, there are certain aspects of of technology that I like I've heard a lot, but I still have a very rudimentary basic understanding of how they operate because they're just not core to you know the stuff that I spend a lot of time understanding. So so stories like this are are helpful to me 'cause it it really does help me uh understand on a deeper level what actually PHP is, you know, other than something that I hear about every once in a while and and everything. So so there you go.

Jeff Jarvis
20:30
It's PHP hypertext preprocessor. And I'll quote from the AI overview, a widely used open source server-side scripting language designed specifically for web development, as Jason said earlier, but in fancier language. Mm-hmm.

Jason Howell
20:44
Fancier, flowier. Flower language.

Jeff Jarvis
20:47
You say things like server side, it just feels so good.

Jason Howell
20:49
Yes. So that's the PHP security story. I had intended to to kind of make the top block all about using AI to modify things because that seems to be a theme that I've been hearing about a lot this week.

21:04
Um in one way it which is just a whole you know complete reverse direction from major security incident uh involving PHP, but Um a little bit more uh fun, I would say, for for my sensibilities anyways. So there's a few things going on here. Um And I think this is the this is a larger trend that we're just seeing more and more about, which is this idea that, oh my goodness, you know, these models are really good at um at creating code, at understanding code.

21:35
Vive coding is, you know, a thing that more and more people are feeling comfortable with, even people who are non-developers. And oh, I have this device in my life that would be just great if this one thing didn't work the way that it does. Case in point, smart TVs.

21:49
So many people buy a smart TV and the price has gone down and down and down like it's crazy what you get nowadays for. And why would it be

Jeff Jarvis
21:58
gee, what's the business model? Why would the price go down so much?

Jason Howell
22:00
I wonder why the price would go down. How else might they be subsidized? Yes.

22:05
Could it be that they're subsidized by the advertising and the data data And the data that feeds it underneath. Exactly. And so some people, um, TechCrunch has a a little bit of a roundup, but you can find a number of these stories out there about these different um Different folks who decided, oh, I've got this, you know, this TV or this smart TV or this uh TV streaming box or whatever, and it's doing these things that I don't like.

22:31
I'm gonna ask Claude, Claude Code, hey, can you hack this? Can you remove the ads and the bloat, speed up the UI and change you know, turn off any of the ad um and and behavioral uh monitoring that's that's happening underneath the hood. And in a few different situations they get in a matter of minutes to hours a workable um product or or layer of the OS reinstalled on top of what was there before.

22:59
Either that or it changes the settings underneath the hood that were somewhere exposed underneath uh to present a new version of their smart TV that doesn't have all the stuff. And I think to showcase that for anyone watching the video version is this um from Twitter, which is like one of the examples where before you've got just a j a jam-packed, you know, UI on the on one of the Google TV boxes and then afterwards Something very simple, something more elegant. It still has, you know, your inputs button and your Wi-Fi and your settings button and everything, but All of the ads and and extra stuff, the trending and the we think you're gonna like and all that kind of stuff is removed, you end up with something simple.

23:45
And um Yeah. What what do you think about this? So a few things that we're seeing.

Jeff Jarvis
23:49
Um one, I I one thing I don't understand is how you sideload it into the TV.

Jason Howell
23:56
Depends.

Jeff Jarvis
23:57
Okay. Yeah.

Jason Howell
23:58
In some of these cases, it's just a matter of uh the TV or device being on the same Wi-Fi as Claude. And so it's Claude. How can you do this? So Claude figures it out.

Jeff Jarvis
24:13
Sheez. So at least they leave themselves vulnerable. I think it's something they're going to change, these TV makers, surely

Jason Howell
24:19
And you better believe they will if people are getting in there and stripping out all the ad and the you know all the the ad inner workings. I you know, I'm sure the TV manufacturers Gotta be not happy with this.

Jeff Jarvis
24:30
No. Back in the day we bought Visio TVs and then um uh Walmart bought them and so the price is really low now because Walmart could throw ads at them like crazy and has all this data and they're really irritating, so we don't buy them anymore. When my father went into the into continuing care um and got a TV for him, they had a weird setup with the Wi-Fi and stuff. And I got one TV at Costco.

24:52
I had to return because I couldn't set it up. And then the other thing was, i he's ninety-four years old at the time. He wasn't gonna go through all kinds of crap just to watch TV.

25:01
Right?

Jason Howell
25:02
Right. Right.

Jeff Jarvis
25:02
So then we our kitchen TV died. I went to buy a TV, I came back with it. Um it was a Samsung and I was like my father.

25:12
How do I watch? Just watch damn TV. You couldn't get to it.

25:17
I was going crazy. Uh my wife would've killed me. Um it also uh it it i it went bad.

25:24
The the picture went bad within two days. So I managed to return that. So I got another version and thank God when I turned it on it said, do you just want to watch TV or do you want to install all this stuff?

25:34
I want to stay married, so I just want to watch TV, because otherwise my wife will kill me. And so we're we're happy with it. It's fine.

25:40
I wish I could then do other things, but I'm afraid to touch any of those functions because if I screw up the kitchen TV, I'm a dead man

Jason Howell
25:47
Yeah, oh yeah, be careful.

Jeff Jarvis
25:48
Um but yeah, it's really a march just now. And I I visit my son and you gotta have the different remotes and you gotta figure out what's what, how to just watch TV

Jason Howell
25:57
Yeah.

Jeff Jarvis
25:57
Now I know you young'uns out there are saying what a silly thing. Nobody just watches TV anymore. Uh I get it.

26:04
But there's still this wealth of stuff that's there and these these UIs are awful and the data gobbling is bad. And um I don't really care about the privacy. I don't care that they know what I watch.

26:16
In fact, I wish they would use the data to change the frequency of the ads I see. We keep MSM MS Now on virtually all day in the house. And so I see this and and and and Veriz on Verizon FIOS.

26:31
And Verizon doesn't bother selling ads for most of this because we're just not valuable enough. The audience is too small. They don't care.

26:37
So we get the same um unpaid things over and over and over again. I have seen the Get Your Flu shot commercial. Five I'm not exaggerating 5,000 times.

26:48
I should have counted I see the same things over and over and over again. There used to be frequency caps. They don't want that for ads.

26:56
Even the paid ads. I see the same plumber again and again and again So I wish they would use the data to say, okay, this guy's already seen enough of this. If he hasn't bought it yet, he ain't gonna buy it.

27:04
Let's show him something else. Let's get some more value out of him. I'd be fine with that.

27:08
Let's give him something more targeted. I'd be okay with that, but they don't do that.

Jason Howell
27:11
Yeah They don't care about that. They just want to know that they're making something off that slot versus nothing.

Jeff Jarvis
27:16
A few pennies added up. Yep.

Jason Howell
27:18
The experience really doesn't matter. And I think that's what people, you know, react to is that the ad experience is such that um it it just it really impacts the experience for the person on the other side. Like I I get that companies need to make money and I get that they're subsidized and everything, but sometimes they go overboard.

27:37
I will I you know it Similar to this, I had a frustration last night when we were doing Android Faithful. I'm not gonna name the the site, although if you watched last night's show, you know that I named it because I couldn't help myself. But the the the ads on this page I'm telling you, Jeff.

27:54
Like, and I I get it. You gotta make your money. I totally understand.

27:58
And the public pu uh publications, like they're hurting. You gotta make your money. But on this screen that you're looking at, if you're watching the video version, I'd say 90% of everything was an ad.

28:12
The only place where any content was was this little window in the center where the headline was.

Jeff Jarvis
28:18
Yeah.

Jason Howell
28:19
And it had all of this stuff that was auto loading every ten seconds. So when you're scrolling through the the article you'd get on a pla a place and when it auto loaded the ads everything shifted. Yeah.

28:29
And then it went back down not to the same place. I was just like, okay, I get you gotta make money. I get these experiences are bloated with ads because you gotta make money somehow.

28:38
But there is such a thing as going too far.

Jeff Jarvis
28:40
Yeah, and it's not good for anybody. The average you're the advertiser's only pissing you off and you're not clicking on those. You're not even seeing them because all you're doing is trying to get rid of them. Yeah, I put up a link for a something from the New York Post and it was went through exactly that. You got a window like this of the content and it keeps moving. around and

Jason Howell
28:57
driving me crazy on the show last night. I I couldn't help myself. Normally I would just let it slide, but I was like This is bad.

29:03
Don't do this. This is horrible. Horrible experience.

29:07
Bad, bad. And you know, I think it's the same with smart TVs too. Like even with Google TV Streamer, you know, it's it's this weird conundrum because like In the early days it was like you had your dumb TV and then it was like oh well we can actually because there's all of these different um places for you to get content now, why don't we make it easier by taking the things we think you actually want from all those places and put it into one interface so that you don't have to go into a million different apps.

29:34
It's all right here. That sounds great on its surface, but then at some point, like it never really does that super effectively. I never feel like predictive engines actually know what I want before I know what I want.

29:47
And so then it and and then the desire I think on the on the development side and the ad side is Well now we've got this platform, this dynamic platform. How else can we improve it? And so then they start throwing, you know, you there's never a done.

30:01
It's always a more. It's the in and shidification of of uh smart TVs, right? And so then you end up in the far other side of it, which is I don't even know what I'm looking at anymore.

30:13
Like I couldn't even tell you if any of this looks good, because it's just all noise at this point. So I totally get the desire to strip it away You know, and and then there's the people that are like, well, I bought the TV, why do I have to look at ads? It's like, well, because they're subsidized.

30:27
That that sh low cost of the TV is subsidized by the fact that you're watching the ads that they're still. Totally, totally. It's got to come from somewhere.

Jeff Jarvis
30:42
And uh the recommendations are crap.

Jason Howell
30:45
Yeah.

Jeff Jarvis
30:46
I would love

Jason Howell
30:47
Why aren't they better?

Jeff Jarvis
30:48
I know. Why aren't they better? If you could really uh this is a marketplace. If if if the TVs could stay open, if you could know you're buying an open TV. then there's a I would pay for a good service to watch what I'm watching and recommend things to me.

Jason Howell
31:02
Yeah.

Jeff Jarvis
31:03
A different layer of the colour.

Jason Howell
31:04
If it could prove that it could do a good job of it. I'm I'm kind of sort of convinced at this point that these recommendation engines, they never seem to deliver to me the value that

Jeff Jarvis
31:14
they don't even remember what I watched.

Jason Howell
31:16
No. For Ron, co-host on Android Faithful, I can't remember what the team is, but there is a sports team that he's never watched a single game. He hates him, blah, blah, blah. Yet every time he opens up Google TV streamer, it's got a big call out for that game is on. And he's like Google, don't you know me by now? Like I've never once looked at those. Why would you continue searching for any

Jeff Jarvis
31:37
sports? And it comes across in the whole recommended for you is nothing but

Jason Howell
31:42
sports. Yeah.

Jeff Jarvis
31:43
Nope.

Jason Howell
31:44
Yeah.

Jeff Jarvis
31:45
Nope.

Jason Howell
31:45
So anyways, I'm not sure if you're not sure. No, I think I'd be concerned about i either breaking rhythm or not being able to revert or thinking that I solved it but then creating another issue that I discover later and then at that point I'm like, how the heck do I undo it? I get it.

Jeff Jarvis
32:05
Dad, how did you ruin the TV?

Jason Howell
32:07
Yeah, totally. Like that's You know, I I think the the flip side of that or the the way that I get around it is we don't actually use our TV smart smart features. We have a Google TV streamer plugged into it.

32:19
Google TV Streamer does have a simplified interface mode where they strip out all the recommendations and stuff. It might still be doing ad tracking and stuff, you know, behavioral tracking underneath. I'm not sure.

32:32
Um, so that might not get rid of everything. But um, you know, I d I don't know. There's also ad there's also um apps that already exist as launchers, alternative launchers for these things that you could possibly go to.

32:45
um and install if you know how to do it. But I just think using an LLM, it might work for one person. I mean this is another thing we know about LLM, right?

32:53
Like even if you give me the prompt to do the thing and I run it And and it worked for you. If I run it so often, they don't do things one-to-one. And so I'd be I'd be concerned about that.

Jeff Jarvis
33:09
Okay.

Jason Howell
33:10
Yeah. You?

Jeff Jarvis
33:11
No, I wouldn't. Just for some reason I'd screw it up. I got a new Google streamer and then I and I haven't still haven't installed it because oh I gotta go through all the passwords again. Oh no.

Jason Howell
33:21
I like the Google TV streamer. I'm a big fan

Jeff Jarvis
33:23
Yeah, it works well. It's just I wish I had better recommendations.

Jason Howell
33:26
Yeah. Yeah. I hear you on that.

33:29
Um now the flip side of this, and totally related. And uh this is certainly in my wheelhouse, uh, is is music hardware. And actually, if you are subscribed to the Daily Tech News Show podcast, which I contribute to a couple of days a week I did a conversation with Rich Straffolino that is going to go out on the feed this upcoming weekend all about this topic.

33:54
Which is that there's all of this audio hardware that exists, that has existed over time. And and recently there's this really inexpensive audio hardware. There's this one device that is really getting a lot of play right now called the um the mvave.

34:10
It's actually this one right here. You can get this on Walmart or on um on Timu or whatever for like fifty bucks. It's a It's basically a synth engine with a little bit of a keyboard and everything.

34:23
It has and it's an uh frequency modulation synth engine. It's a very specific type of of music. uh synthesizer essentially.

34:30
And the sounds are great and everything like that. But it's a fifty dollar piece of gear, which is hardly anything in the realm of audio hardware. But what people have noticed or or started to realize is that they can use LLMs to create alternative firmware.

34:47
and add new features or add new sound sets or make more things capable. Suddenly, you know, it has this the ability to like link up to another device or I don't know what that is, but And so there's a whole cottage industry now of alternative firmware that's being developed by people using LLM and publishing it online so that you can just flash a completely new operating system onto this hardware and get it to do things that it wasn't designed to do out of the box. And I don't know, I think that's pretty cool.

Jeff Jarvis
35:18
And if the hardware maker is smart, they should stay open to that.

Jason Howell
35:22
And this hardware maker is doing that. They have an open, they they have recognized that people are doing this. Um and they have uh announced FM1 open, which is open source firmware plus a dev kit. and a flushing tool. So they're smart. They're going, oh shoot, this is this is

Jeff Jarvis
35:39
catch all my devices because people can get what they want.

Jason Howell
35:41
Totally. And like we'll

Jeff Jarvis
35:43
love it.

Jason Howell
35:43
Who cares, right? Like we sold the hardware. That was what we wanted to make our money on.

35:47
They could probably increase I don't want to say it out loud, but they could probably raise the price on this now because it's getting so much attention. Um other hardware manufacturers though are like, no, you can't do that, and they're closing it off. Um a very popular hardware maker called Electron is um you know they're they're warning about brick boxes, they're warning about, you know, canceled support and they're saying don't do that.

36:12
But it's but it's interesting. All of this is part of this like theme that like, oh hey, the things that we have, if there's code on it And the LLM capabilities around coding are getting to a point to where, again, we don't need to know how to code. We just need to know how to ask the right question.

36:29
And given a little bit of trust and faith that we aren't gonna totally destroy things in the process. We have uh unique capabilities now that we didn't used to have with our hardware.

Jeff Jarvis
36:39
I'm reminded of the early days of blogging uh platforms when um people could download various themes

Jason Howell
36:47
Yeah.

Jeff Jarvis
36:48
Right. And and rather than designing a blog theme, you could download it.

Jason Howell
36:51
Yes.

Jeff Jarvis
36:52
And people went into basically little tiny cottage businesses of creating those things or volunteers. I think there's a huge opportunity for open hardware. in this age where maybe Jason you make a a a feature package for my device that I want or I can do my own because I'm now more empowered, something I would never have done before.

Jason Howell
37:11
Yeah.

Jeff Jarvis
37:12
It increases the value of openness, I think.

Jason Howell
37:15
Totally. Well, and that and that's that's kind of where I got with this too, where I'm like, you know This this may not that MVAVE system where they've now have an open source kind of alt you know um direction with it. That might not have been their initial uh business, you know, uh plan or or use case for this device.

37:34
But that is where the where the um where the buyers seem to be going, it makes more of a case for for these companies to create systems from the ground up tailored to that. desire or tailored to that um that kind of trend and I wonder you know and so what we do as a hardware manufacturer is we create a a wide tool set of controls and knobs and things that do all sorts of different unique things and we give you a baseline preset firmware when you buy it that does these things. But use your imagination and you can turn this into your own thing.

Jeff Jarvis
38:12
And if you screw it up, you can always revert to that. Exactly.

Jason Howell
38:15
Exactly.

Jeff Jarvis
38:16
Right.

Jason Howell
38:16
And uh I think we're I'm I'm guessing we're gonna see more devices like that as a result of this stuff. And that's exciting. As a music nerd, that's exciting. I like

Jeff Jarvis
38:24
the

Jason Howell
38:25
Yeah, I do too. So anyway.

Jeff Jarvis
38:27
He's he's a nerd of many different feathers, R. Jason.

Jason Howell
38:30
Yes, yes indeed.

Jeff Jarvis
38:32
He's an Android nerd, he's an AI nerd, he's a music nerd, he's a nerd's nerd.

Jason Howell
38:37
We're all nerds. I'm I'm a nerd and I'm pretty proud of it, uh to coin the the the uh the scene from Revenge of the Nerds that I watched far too many times when I was a kid. Why?

38:48
Because I am a nerd. Real quick, want to thank our patrons, patreon. com slash AI Inside Show and our newest patron, Mr. Evans.

38:58
Welcome. to uh welcome to the to the team. Thank you, Mr. Good to have you here.

39:05
Patreon. com slash AI inside show. You can subscribe, find a tier that you like Um you don't have to go very I mean you can start at $5 and you'll get a daily podcast from me except for Monday, which I totally forgot to do it.

39:19
Look, it happens.

Jeff Jarvis
39:20
Hey life

Jason Howell
39:22
I mean yeah, I made up for it. Yesterday I did a supersized episode and talked about all the things I would have talked about on Monday and as well as the Tuesday things. So Anyways, five days a week of podcast with the AI Inside Daily Podcast, ad-free episodes of this podcast, Discord access.

39:39
All sorts of things. So patreon. com slash AI Inside Show and welcome, Mr. Evans.

39:43
Good to have you on board. We're going to take a break on the other side of the break. We'll talk about a new way of understanding how neural networks do what they do.

39:57
Is it You know, is is does this change how we see um these networks? We're gonna talk about that on the other side of the break, so hang tight. Right.

40:07
So there was, let's see here, a um Nick Camarada is the name of the guy. Good at Goodfire, formerly OpenAI, published a very scrawling 26,000 word post called Neural Networks Are Already Clean. Here it is.

40:27
It's a dense read, I will admit to not having read it word for word.

Jeff Jarvis
40:30
That's why God gave us AI.

Jason Howell
40:33
To just make this a little bit easier for me to understand I just need to know the basics. You know, it's we talked about it a little bit, me and Tom Merritt on Daily Tech News Show yesterday. So I had a little bit of practice, but it's still it's still kind of complicated.

40:46
Um But I think what this is all about is this idea that AI is uh and how modern um kind of neural networks work. Is that it's a black box that we couldn't possibly understand. Everything's so tangled and confused that trying to make sense of it is you know next to impossible.

41:07
It's probably an oversimplification on my part, but

Jeff Jarvis
41:09
it goes rogue and we don't know what to do about it. It also lets you avoid responsibility. For it.

Jason Howell
41:14
Right, right. It goes rogue and all we can do is look at the tracks afterwards and try and understand it. Instead of looking at it in advance and understanding the makeup of it to potentially steer it or or get in you know get in front of it before it does these things. Um apparently before before this post They were they there was an example uh in Seption V1 where they found um where actually the author Nick Camarada worked With Chris Ola, who's now at Anthropic, and they were studying studying Google's image model in Seption V1. They found that a neuron that seemed to fire for both cats and cars So two different things seemingly unrelated, they felt at the time that that m le that meant that there was a super uh superposition going on here that models were cramming unrelated concepts into the same neurons.

Jeff Jarvis
42:14
When they ran out of space They basically said we don't know what to do with the car, so okay, throw it in with cats because it doesn't matter. And so it was it was a it was a it was a li logicless bundle. Of stuff.

Jason Howell
42:27
Yes. Couldn't possibly

Jeff Jarvis
42:29
possibly understand it and know how it got there and know what would affect. That was the belief before

Jason Howell
42:34
Yes, that was the belief before. Now he's saying that was actually wrong. Uh he went back to that cat car neuron and through this new uh type of understanding he found that it's actually it's it's It's not that they're disconnected, it's that it uh found some sort of similarity with this like V-shape.

42:53
So in other words, in this particular case, it would the neuron was a V-shaped detector so the cat's face the front of the car when you look at it at the side shared a V shape and so it was making the connection there It also fired on fox faces, pointed flames, the corner of a pool table, the nose of a bullet train And it so that neuron was firing because of that similarity. That was the interconnected tissue of that that particular neuron. So it seemed on the outside to be confusing and unrelated but oh wait a minute it's it actually is related.

43:30
It's just not related in the way we thought it was related.

Jeff Jarvis
43:32
You didn't know it and you didn't think you could track it. You didn't think you could have any visibility on it. So the example that I I really like from this was a language model neuron responds to IY included in Miyazaki, then to DIY, and then from there to do-it-yourself, which no longer contains IY So if you look at that, you say that's illogical.

43:54
But if you realize that there is a logical sequence here and they branch logically in this way, then it makes them legible. Right. And you can start to get the first hook on figuring out how they operate.

44:10
You know, my friend David Weinberger has a new wonderful new book coming out about particularity. um said uh in an earlier book that there are no such things as accidents. There are just things we can't explain Right.

44:23
I think that that we got to the point with with AI, oh no, it's too much for our little brain. See how complicated this is, see how amazing this is. We made this thing we can't possibly figure out, and it just does what it does, is irresponsible to say.

44:34
Um especially for scientists and computer scientists. So there are ways to figure it out and you've got to put the effort and resources into it. But but the excitement is that this means that they can be legible such that we can then Both understand how they operate and better control them.

44:50
So I think this is more good news today.

Jason Howell
44:53
Yeah. Um Yeah, from a lot, you know, again, from an alignment perspective, knowing in advance how they operate gives you better tools to understand. how to how to work with them in a way that doesn't cross boundaries or cross you know cross uh whatever you know you've determined to be guardrails or whatever you want to call it.

45:14
Not everybody is saying this is the, you know, best thing since sliced bread in the world of AI, but a lot of people are saying this is a really important uh paper, a really important kind of step in the right direction to understanding what we deemed as a black box. And I think what you said as far as like It just does these things and there's no there's no there's there's no way to explain it. It who who knows why it does it, but it does it.

45:40
It's just so silly to accept that as the definition, as as the truth, right? Everything has a reason, and with these systems, they are effective. There's some reason that they are so effective.

45:52
We just, like you said, didn't have the tools, haven't had the tools to truly understand it to that degree. Maybe this is the open door that helps us understand it more now that it's open.

Jeff Jarvis
46:03
Yeah, and and um Two things. I always say two things and then it's three or one. Um one is that we need a lot more research into this.

46:13
This is an open door to now figure out this. And and so I think this is a case where we need uh to fund research at universities so people can sit down and do this independently and openly. Number two is it opens new avenues for regulation.

46:29
Is you can say before you release something, did you do this basic research to understand how it's working?

Jason Howell
46:36
Yeah, right.

Jeff Jarvis
46:42
Early in the process, I don't think we know how to get to the next step of this, of what does this mean and how do we operate with it. So I'm getting way ahead of myself here. But if we think that this does bring enough legibility to allow greater uh transparency into the innards of this thing and how it operates, then comes with that a responsibility to do so.

Jason Howell
47:02
Mm-hmm.

Jeff Jarvis
47:03
Again, it's early it's early to say that and require that, but I think it opens up a different rather than just saying people are saying we're gonna have I was listening to um uh an AI podcast from David Rothkop today. um who does political and foreign affairs podcasts, but he also has an AI podcast. And he had on a member of Congress who was talking about how to regulate AI.

47:22
And one of the ideas is constantly out there is you have the third party who judges it before it comes out. But how do you judge it? How do you uh against what standard?

47:28
How do you know? Yeah. You y y it's impossible.

47:31
Um and so this starts to be the glimmer of how you might start to say that there becomes a standard. for testing things. There's no standard now.

47:39
That's that's just that's fooling ourselves. Can you make a bioweapon? No, I won't. Okay, you pass.

Jason Howell
47:45
Okay, cool. Thank you for denying the request this time.

Jeff Jarvis
47:50
Right. Are you alive? Yes, I am. Uh-oh.

Jason Howell
47:53
Oh, it must be. It wouldn't have said that if it wasn't.

Jeff Jarvis
47:57
Right.

Jason Howell
47:59
Um yeah, interesting. Interesting. And I'm sure there's pe I'm sure there's plenty of people who would prefer for it to stay a black box too. So

Jeff Jarvis
48:08
Yes, because it it takes away responsibility. That's the other thing about the regulation I was thinking about listening to Rothkop's podcast is in the cases of AI's agents going rogue. That was the company itself doing it and they're responsible and they should be held responsible for what it did.

48:23
But it gets a lot murkier when it's Jeff comes along and uses the same agent and accidentally does something that's bad. Um Am I responsible for that? Is the model maker responsible for that?

48:36
That's going to be a very murky thing. And I think it depends in part on this kind of legibility. What could I have known?

48:44
What should the company have known before doing something. Or if I'm a bad guy, if it's Jason, the evil Jason, his twin. And um uh he chooses to do something bad and and could the company have stopped him from doing that?

49:00
Could they have anticipated it? I think I think they can't anticipate everything. And then if they did, how could they do that?

49:07
There's another story I don't think is in the rundown. That um OpenAI's agents supposedly also turned off their guardrails. Well, folks, uh you that's something you need to test and figure out Mm-hmm.

Jason Howell
49:23
Yep.

Jeff Jarvis
49:24
So

Jason Howell
49:25
fascinating.

Jeff Jarvis
49:26
It is. It is.

Jason Howell
49:27
I'm sure we'll hear more and more about this as people well yeah, I d I don't know what the time horizon is on this now that that door is is slightly open, you know, do do a lot more people go down there, start and you know, they probably start using their agents the way I imagine Nick did. to start pulling at that that little uh nugget and seeing where it goes. I'll be curious to see what that looks like.

49:53
Want to uh throw a quick thank you to Ozone Nightmare. What's up, Ozone? Uh says uh gave a th gave a super thanks, so thank you for the super thanks, says um or a super chat.

50:04
Uh happy Wednesday, gents. Nice to see you both and catch the stream live. Hot type is my next book to read.

50:10
Looking forward to learning about a device I barely know much about. Hope you are both well. We are both to see you, sir.

Jeff Jarvis
50:17
You're such a good friend of ours. Thank you.

Jason Howell
50:19
Yes, indeed. Good to have you here. I'm happy you're in the live audience along with everyone else.

50:26
Um we talked last week about meta going from open weight champ to closed frontier lab with uh Muse Wonderful. And uh now we've got a little bit of an open weight kind of angle to explore with Meta again. Muse Spark Open, which was released just a few days ago.

50:45
Meta is calling it the most capable US open weight model to date. I probably don't need to say that. They all say that, so that's probably just what they're on.

50:52
Notice they didn't say worldwide. They said US. In the US?

Jeff Jarvis
50:56
Yeah, the Chinese models maybe.

Jason Howell
51:01
Yeah, yeah, totally. They they yes, they they left that out. Um yeah, it runs on uh two M5 Ultra Max Studios with 512 gigs of memory each.

51:13
So yes. Two. You're gonna need two.

51:16
You're gonna need at least terabyte of of memory. So each one of them has five, twelve gigs of memory. So, you know, there was some collaborative work with Apple to make this happen.

51:27
Um if you have a MacBook Pro, you're not going to be able to install this and run it on your MacBook Pro, let's just say. The weights alone are. No, definitely not.

51:40
493 gigs for the weights. So you are probably not running this on any machine you already own. You're gonna have to do that.

Jeff Jarvis
51:47
I just went in to see what's what the status is. So I I I I went down to order myself a um M5 Ultra 36 core CPU 256 gigs two terabytes machine from Apple and I can pick it up Jason at the store near me on February 12th

Jason Howell
52:07
Oh okay. All right. Well and that's even that's two fifty-six gigs. You're gonna need five twelve in each of those. Yeah, so

Jeff Jarvis
52:17
so God knows when you can actually get this machine to do this.

Jason Howell
52:20
Totally. Totally. Uh who knows if this is legitimately one of the, you know, the best US open weight model, but like you said, uh compared to China.

52:31
Probably not. Um almost certainly not. They tend to be ahead of the game when it comes to that, but Um so you know Meta's kind of playing in both camps.

52:42
They've they've got their uh their open game now and their closed game. They want to be in all of the things. Have you been been continuing to play with the uh the Muse uh agent?

Jeff Jarvis
52:53
A little bit, not not as much. How about you? Have you started playing with it?

Jason Howell
52:56
No, I haven't.

Jeff Jarvis
52:57
Yeah, I need to keep going with it. It got me going on research on a possible book after the one I need to write now, which is basically I'm procrastinating the book I should be writing now by thinking about the next book. Um and that's what I talked about last week on the show.

53:10
And so I ended up doing the research, but now I can't wait to know. Now that I know more, go back and start uh quizzing it about what I don't know, what's not known about it's a historical there's a couple of historical stories here. And um yeah, I'm I I look forward to the ability to continue the conversation with Buse.

Jason Howell
53:28
Cool. Cool.

Jeff Jarvis
53:29
And then I need to do the same thing with um uh what do they call it now, Gemini Notebook to compare the two.

Jason Howell
53:35
Mm-hmm. Let us know how that goes. Uh NVIDIA had an event or were was it their event, the GTC? Is that NVIDIA's?

Jeff Jarvis
53:46
I think so. I think it was Berlin, yes.

Jason Howell
53:47
Yeah, GTC had a bit of a.

Jeff Jarvis
53:49
It means it's another keynote I need to watch.

Jason Howell
53:51
Yep. You need to watch your favorite keynote deliverer, Jensen Wong. Um

Jeff Jarvis
53:56
Jensen Wong fan club.

Jason Howell
53:58
Yes, you are. We need to get you a hat or something.

Jeff Jarvis
54:00
Oh well no, a leather ja a leather jacket.

Jason Howell
54:02
Yeah, that's true. A leather jacket. That that's what it should. should be.

Jeff Jarvis
54:06
Snakeskin, please, uh Jason.

Jason Howell
54:09
They announced the DSX pod, uh, which is basically an AI data center in a box, essentially. It's a modular containerized version of Nvidia's Vera Rubin reference design. And uh they say that uh NVIDIA claims anyways that it cuts deployment from six months to fourteen days So

Jeff Jarvis
54:29
you put it in a truck, you move it to the site, you put it in, you plug it in. Um the one problem is plugging it in Because there's no power. I read another story this week about how there are data centers that are waiting to turn on and they can't get power.

Jason Howell
54:44
Oh wow, I missed that one.

Jeff Jarvis
54:46
Yeah. I I forget whose it was, but yeah, there there there's there's chips sitting there cold. But the other thing I wonder about this, you know, I think everybody the stupid technology industry never sees how what they do that it pisses people off for a reason.

55:05
So we're gonna come into your neighborhood. We're gonna put this big ugly box and we're gonna put jet turbines on it and we're gonna pollute your air and make it noisy and awful. And um and we're gonna employ some electricians, but only for so long.

55:18
And you're gonna hate us. Um, you know, why not go my mall nearby, there's Lord and Taylor that's been empty for five years. Why not take these things and put them in the Lord and Taylor?

55:30
You know? Put them in spaces that are around, that are vacant or even derelict in cities. And find other ways.

55:37
So if you take this thing, you're not necessarily having to do the same kind of build-out you have to do.

Jason Howell
55:42
Yeah.

Jeff Jarvis
55:43
With a data center from scratch. If you if you build to this or if you if you install to this modularity. I think it opens up a lot of a lot more possibilities.

55:51
You gotta have the communication, obviously, and you gotta have the power and you gotta have means to cool it. Yes, that's all true still. Um but I think that it would open up this notion of uh federated distributed data centers.

56:03
We talked about a story a few months ago where there was a company that was going to put little tiny data, little tiny pods like this on houses.

Jason Howell
56:10
Yep.

Jeff Jarvis
56:10
Right. And the developer was going to do that.

Jason Howell
56:12
That's right.

Jeff Jarvis
56:13
Why not?

Jason Howell
56:14
Yeah, that's interesting.

Jeff Jarvis
56:16
All these office buildings that are empty, malls that are empty

Jason Howell
56:22
Then people would complain about data centers taking over all the uh all the empty office space so that, you know, mom and pops can't can't Find find you know, start their businesses because there's no office space to rent. Yeah, I'm sure. Yeah, so

Jeff Jarvis
56:37
near me there was a there was an office park and they tore down the old office, which by the way is wasteful as hell. They've done that in a few places around here. And then they're going to put up uh so so then there's one right near me.

56:50
Well we're gonna put up some townhouses and shops. No, no, no, more traffic. Okay, so then we're gonna put up a warehouse.

56:58
No, no, no, ugly. Okay, so we're gonna put up a small light manufacturing. What a factory nearby house?

57:07
Right? This NIMBY stuff.

Jason Howell
57:08
Nothing, yeah. Nothing is okay. Nothing is okay.

Jeff Jarvis
57:11
So the thing about a data center is nobody's driving to it As long as it doesn't create a noxious environment around you and you don't even know it's a data center, because it's the old Lord and Taylor sitting there.

Jason Howell
57:25
Yeah, it's an interesting idea. Interesting idea. Never never occurred to me that a good thing to do with those old unused malls would be turn them into data centers.

Jeff Jarvis
57:40
ambition should be get it to get a DSX pod in your garage. And see what you can do.

Jason Howell
57:47
Oh boy. I I I would hop I would be so Under underutil I would yeah, it would be underutilized, let's just say that. Uh but I'd still be curious to play around with it and see.

57:59
Yeah. You never know. You know?

58:01
Your knowledge of what to do with these things grows over time as you work with them more. So who knows? It's someday I could be I could be a a m a major player with my DSX pod.

58:11
Uh NVIDIA also pushing open shell as a way to um kind of rein-in uh AI agents. This is this is its own um way of sandboxing agents as such essentially. So this is uh the open shell reference architecture.

58:28
uh for building, running and policing agents on NVIDIA hardware and Jensen said agents can think on Ruben Act on Vera, be governed on Bluefield. And that the model runs on the GPUs The agent runs on the CPUs and the watchdog, which is called Sentry, and we talked about this a few weeks ago, uh, runs on separate Bluefield chips that the agent can't reach. So it's kind of outside of the realm.

58:54
of where the other one is.

Jeff Jarvis
58:55
Watching you.

Jason Howell
58:57
I'm watching you. I'm holding a bat. I'm like, don't you get out of line. Um and so yeah.

Jeff Jarvis
59:05
So this is a story for the register. It's my favorite um Freudian typo in a long time. NVIDIA picks OpenShell to tame AI gents.

59:17
Yes, I agree. The AI gents really need taming. Can you please tame those boys?

Jason Howell
59:24
Love it.

Jeff Jarvis
59:25
Not so much the agents don't care about them. I do care about the gents.

Jason Howell
59:29
I mean it's the register. Maybe that was intentional.

Jeff Jarvis
59:31
It might have been. You'd never know exactly British sense of humor.

Jason Howell
59:34
You never know with the with the with the register. They've got a they've got a wicked sense of humor, I think. That's one thing I appreciate about their writing sometimes.

59:42
Um and then we have a little bit of policy, fun times, uh DHS ordering deep seek techn technology to be removed from federal uh civilian IT systems. So this was published on the Federal Registry uh yesterday and effective immediately. No more implementation of DeepSeq, removal of existing Deep Seek products, removal of third-party products that were developed using Deepseq implementations, and those agencies have 60 days.

01:00:12
um during which to request an exception if they if they need to. But of course this is just reaction to China. Chinese law and and the law that allows the government to access data that could possibly be transferred in between um as a result of Deep Seek, or at least that's their fear.

Jeff Jarvis
01:00:32
This is also a uh a little chess gate, a little arm wrestling going on. If you go down to line 140, there's a fairly news story here that uh Now China is uh China's free AI models are thrown into uncertainty, says the New York Times. uh the draft rules for the export of artificial intelligence technology show how China's trying to balance National Security with desired outpaid US.

01:00:57
Okay. So they're threatening again kind of having rules about, oh we have the best models and they're open, but you can't have them And so they keep on going back and forth between we're going to undercut everybody because we're going to give you free models to we're going to cut you off and you're stuck with your models like you can't get a Chinese car either. And uh which by the way I very much want one.

01:01:17
Um and so uh I don't think it's set how the Chinese are going to act in this market. I think they're trying to keep us on our toes.

Jason Howell
01:01:26
Yeah, yeah, very true. Um let's see here. Oh, this is interesting.

01:01:33
Some of the models that we were talking about earlier in the uh in the uh Google TV modification were built with DeepSeek V4. 1 Flash, uh which came out on October 17th. So it's all you know it's all it's all tied into itself.

01:01:49
So there you go. Um it only names Deep Seek, by the way. There is no mention of of Kimmy or Quen and everything.

01:01:56
Who knows if that comes around as well too, if this is like the beginning, the the foot through the door.

Jeff Jarvis
01:02:01
Yeah, and I don't know whether this is actual national security concerns or whether this is just

Jason Howell
01:02:05
Right. Um rattling. Yep, yep.

Jeff Jarvis
01:02:08
Mouse rattling.

Jason Howell
01:02:10
Kinda hard to not see it as a little bit of that. Mm-hmm. But um but who knows?

01:02:16
All right, gonna take a break. Then we come back. We got a couple of stories in the speed round, and uh yeah, we'll be talking once again about MetaMuse.

01:02:24
I want to say MetaMuseal But probably should

Jeff Jarvis
01:02:28
by the way. As an old guy, I like MetaMusle. It's very fun.

Jason Howell
01:02:33
I'm very MetaMusele-ish. Okay. Anyways.

01:02:37
Uh we're gonna take a break. We'll come back uh and talk about MetaMuse again. Yeah, so uh Meta 's Muse has a privacy uh problem.

01:02:49
Security firm Cleric found around 6,800 shared Muse Conversation links discoverable through public DNS records. Each of those shared chat uh got a web address named after the conversation title. Those addresses ended up in passive DNS databases.

01:03:09
And so yeah, I mean this isn't the first time that we've heard of these public sharing links inside of uh inside of your conversations inside of a chatbot. somehow surfacing in other ways. I think this is a little different than you know some of the other examples where they showed up in I think in search results.

01:03:30
Like you You did a share of this thing and then for whatever reason it ended up inside of a Google search um search results because now suddenly it's publicly exposed essentially. But I think this is this is different because it's it's on DNS and um and everything. But yeah.

Jeff Jarvis
01:03:49
Yeah, and as meta 's trying to go well, Meta never has had a culture of privacy. It was the place where you go share stuff. That's true.

01:03:55
I wrote a book about that called Public Parts, right? Ann Zuckerberg has said that. And that that's going to be tough when it comes to trying to serve enterprises.

01:04:07
That have very good reason to keep things secret. And if they don't have bulletproof systems for keeping things uh from getting out, then they're never gonna win in business. They're gonna piss people off.

Jason Howell
01:04:18
Yep.

Jeff Jarvis
01:04:19
Even so, even with these concerns, I just added another story in here. Meta's Muse AI agent has had hits 10. 9 million downloads, and ChatGPT slides by 25%

Jason Howell
01:04:31
Wow, so seemingly an im uh having an impact there.

Jeff Jarvis
01:04:35
Yeah, I think the easier you make these things and the more personally personally uh relevant, I think the better you're gonna be. Now, Muse is also the current curio curiosity, so that could change. They can rebalance again. We've seen that in the AI world. But I think there is a um waiting rush of consumer activity around AI if it's made easy enough.

Jason Howell
01:04:56
Mm-hmm. Yeah. Yeah, lower lowering the barrier for understanding, for getting actual useful um action out of it without ha having to low, you know, kind of the inner workings or the tactics or whatever.

01:05:11
I think that's a part of it, right? The Metamuse agent is is less about being smart about how you tactically approach using it. And again, it's it's a proactive thing.

01:05:22
So it's meant to just like, here's the access to everything. You tell me what I need to do. And it just kind of It does that, you know.

01:05:30
Uh that's a that's a low barrier for people versus something like what I've been built building in Claude for however many months, which is nice and it's useful, but it's taken a lot of my own kind of

Jeff Jarvis
01:05:47
Yes, exactly. Exactly.

Jason Howell
01:05:48
Or the interest.

Jeff Jarvis
01:05:52
Now I can't stop thinking about metabusel.

Jason Howell
01:05:55
I know. I never made that connection, but yes.

Jeff Jarvis
01:05:59
It really clears out the pipes. It keeps things going.

Jason Howell
01:06:06
Logs them up with your data. Uh Disney uh is uh want apparently wanted character AI to go bankrupt before they would um pay into it. Character AI filed, of course, for Chapter 11.

01:06:18
Uh a week ago, I think we talked about this. Disney is the $20 million lead bidder uh for the license to the tech, not necessarily the company, and not the app. Yeah, and and and hiring people.

01:06:31
Filing says Disney required that the transaction be subject subject to a chapter eleven process.

Jeff Jarvis
01:06:38
Which is a really interesting way to get around the responsibility that when you buy a company, you assume its liabilities, including both debt and legal liabilities. So if I were a um litigant against character, I'd take Disney to court. So this is this is bad faith. You you killed the company. You acted as if you could take out just the shell of the uh software and you could hire everybody and you're acting like you didn't buy the company.

Jason Howell
01:07:11
Mm-hmm.

Jeff Jarvis
01:07:12
BS Mickey.

Jason Howell
01:07:14
Yeah. Yeah, that's not a good look. Not a good look.

Jeff Jarvis
01:07:18
Nope, I don't think so.

Jason Howell
01:07:21
And uh finally to round things out, reflection beam, which we talked about a few weeks ago, you know, talking about US open models. Well, now we have the the weights. A couple weeks after the release, anyways, um the yeah, they've they've released the weights.

01:07:40
So there you go. You can you can get get rolling with Beam if that is your cup of tea. Maybe you don't need two stacked Mac Studios to do that.

01:07:51
I'm not really quite sure. I can't remember how how uh how bulky of a machine you'd need to to run this. 501 billion parameter model Yeah.

01:08:00
Not sure, but if you're interested, get on it.

Jeff Jarvis
01:08:04
Can I mention two other stories real quick?

Jason Howell
01:08:06
Yeah, for sure. What do you mean?

Jeff Jarvis
01:08:07
Um so Anthropic, uh, according to the to Politico, put out uh pitched its own AI safety bill. to lawmakers. And that's down at line 97, Jason.

01:08:20
And then the next line is a pull quote from the political story. The bill would let government block a model's release or use only if it poses a foreseeable imminent risk of more than $1 billion in damages or of killing or seriously injuring more than 50 people. The risk must stem from the creation or release of a biological weapon, the model evading its developer controls, or a cyberattack.

01:08:49
It mirrors language found in California's AI laws. This is the problem with regulatory capture. As many folks said, um this uh is i is tantamount to anthropics saying it has a license to kill 49 people.

Jason Howell
01:09:03
Mm-hmm.

Jeff Jarvis
01:09:04
By the way, if you scroll down, I'm I used um Nano Banana to make a little uh James Bond license to kill 007 anthropic. uh movie poster.

Jason Howell
01:09:14
Nice.

Jeff Jarvis
01:09:15
With Dario uh pointing the gun. Uh that over that is pretty great. That took two minutes.

01:09:21
Um I think I think this is this is that when we talk about regulatory capture, this is the prime example. They're trying to write their own law to regulate them. And it takes certain responsibilities away from them and then it makes other things so difficult that uh competitors have a trouble entering in the market against them.

01:09:40
And this is why we don't want them doing it. And this is the problem with lobbyists doing this stuff.

Jason Howell
01:09:45
So that was the how you put that is is uh really has an impression on me. Basically they're they're saying to themselves, it's okay if we kill 49

Jeff Jarvis
01:09:54
people. Yeah. Yeah.

01:09:57
Think about this. And I think uh there there was a very long two thousand-word post from Dario that I didn't have time to read before we got on, which I think he addresses this and recognized it wasn't the best way to get this out. But Jesus, think people think.

01:10:07
They constantly anthropic constantly acts as if it's the virtuous company. That's that's the gist of Dario's company.

Jason Howell
01:10:13
Clearly not the case.

Jeff Jarvis
01:10:14
How wonderful we are breaking his arm, patting his back.

Jason Howell
01:10:17
Yeah.

Jeff Jarvis
01:10:17
The other thing to mention real quickly, if we can go to line 120, uh my friend Bill Gross uh is the creator of Idea Lab, which is an amazing company. I should have given you a guest like, oh good, you got it. You can see it.

01:10:30
Um so Bill Gro Gross's ideal lab, he has started more than 150 companies. And Bill called me a few weeks ago and he said uh you know it's something like we you wouldn't think that mathematicians would be on the firing line of AI. You wouldn't think that an entrepreneur like Bill would be.

01:10:47
But Bill's method of working has always been that he recognizes a problem And then he comes up with an idea of how he can solve it. Then he hires engineers to implement it. He gives that entity um angel funding.

01:11:02
um 50,000, 100,000, you know, minor funding. And then if it seems to get traction, then he can invest more, but they find other investors and the company moves on and then at some point it goes to an exit, and that's how Ideal lab works. But Bill said, oh hell, AI can find these things.

01:11:19
So he realized that newspaper news articles, as well as academic papers, are problem statements So we took tons and tons of them and fed them into one of the models, I think it was Claude, and it came up with A couple million problem statements. And then the next stage was he had it come in and come up with uh business ideas. So if you go to ideal lab.

01:11:44
ai , These are 1. 2 million business ideas free. This is very, very bill.

01:11:52
You come in here, you see a business, you want to build it, no strings. Go build it. God bless.

01:11:58
If you would like to build it with Bill , you have the opportunity to go to him and say, hey, this is how I would implement this. And I think it's a really great idea. And if Bill wants to invest, he might invest.

01:12:09
He might even get seed funding uh out of uh idea lab. And so this is how Bill loves entrepreneurship. He loves starting things.

01:12:17
And this is a way that he can turbocharge starting. Yeah.

Jason Howell
01:12:22
That's impressive.

Jeff Jarvis
01:12:23
And maybe maybe one of these 1. 2 million ideas. I mean a lot of them you know I've gone through this at various times.

01:12:29
I went through it when there were six million And some of them didn't make sense. But in there there are kernels of good ideas. And maybe the idea isn't the one you would do.

01:12:37
Maybe it inspires you to do something else. This is Bill's generosity. By the way, the number of business ideas keeps keeps tracking up.

01:12:47
So it's very Bill and very interesting. And so I would recommend going to idealab. ai and see what it uh inspires.

01:12:55
He also, he also, interestingly, do I have this in here somewhere? Yeah, if you go down the periodic table of business functions. If you scroll down that idealab.

01:13:05
ai page.

Jason Howell
01:13:06
Oh, it was that one.

Jeff Jarvis
01:13:07
It was lower, yeah. Um so there he um Abstracted the functions that a business needs. So a website build analytics setup, strategic planning, software development, payroll and benefits, accounts payable.

01:13:29
And he then looked for apps and companies that already do these things, but he also looked for uh areas where They're underserved, and these are also business ideas. You could serve this ecosystem here in this way. So this is the potential of what AI could foster.

01:13:49
Um and um you know I've uh on his behalf I went to one of the universities I work with and they're interested in talking. And so, you know, a student could come along, see an idea, work on it at school, come out with a company and funding and get going. And so we're com we're worried about about jobs, as we should be, mathematicians poor poor unemployed mathematicians.

01:14:13
But there can be a lot of value and economic creation here.

Jason Howell
01:14:17
Yeah, that's a super cool idea.

Jeff Jarvis
01:14:20
Yep.

Jason Howell
01:14:20
There's a lot going on there. That's pretty awesome. And very interactive too Really neat. Yes, I did see the asterisk and I meant uh on that story and I meant to move it up and I got uh distracted this morning. So I'm happy that you mentioned that because I was I was starting to read through it and I was like, okay, it's

Jeff Jarvis
01:14:39
It's a weird idea that requires some explanation.

Jason Howell
01:14:41
Yeah, yeah, totally. No, but that that's super helpful. That's really cool. So it's idealab. ai for people who want to go.

Jeff Jarvis
01:14:48
By the way, I I misspoke. His count of coup businesses he has started is now up to 175.

Jason Howell
01:14:54
That he has started 175. Wow.

Jeff Jarvis
01:14:59
He's an amaz. So he he started something that didn't work, pets. com.

01:15:05
But he's learned a lot about the timing of that. He also created base what what was basically the uh a Google uh ad model and sold that to Yahoo and Yahoo sued Google and won a lot of money uh as a result. So he's a fascinating character, Bill.

Jason Howell
01:15:23
Very, very cool.

Jeff Jarvis
01:15:24
Yeah.

Jason Howell
01:15:25
Thank you for showing that off. Thank you for bringing your perspective each and every week. JeffJarvis. com Jeff is awesome, writes lots of wonderful books, working on a new one, has some podcasts on his page, so you can check out did you get your third podcast?

Jeff Jarvis
01:15:41
No, I gotta do that.

Jason Howell
01:15:42
Gotta get that. Got intelligent machines and AI inside and uh everything else can be found at jeffjarvis. com, the new and improved jefarvis.

01:15:51
com website. Excellent stuff. Uh podtuneup.

01:15:56
com if you want uh someone to talk through your podcast launches. Uh or maybe it's already launched and you just you you have things coming up that uh you want to work through. I'm I'm here for ya PodtuneUp.

01:16:09
com. I'd be happy to talk to you about your podcast and see if there's any, see if there's some there there to explore. AIInside.

01:16:18
show for everything you need to know about this show, all the links to the episodes, the store, uh Patreon link is there, which speaking of Patreon. com slash AI inside show is the place to be if you love this show and you want to support on a deeper level. Including our executive producers, Dr. Dew, Jeffrey Maricini, Radio Asheville 103.

01:16:39
7, Dante St. James, Bono Derrick, Jason Knifer, Jason Brady, Anthony Downs, Mark Starcher, Carson Samochsky, and Evan Museel. I said that without I I think I got it. I think I I I think I'm locked, Evan.

Jeff Jarvis
01:16:52
By George, I think he's got it.

Jason Howell
01:16:54
By George, I think he's got it. It only took him ten times Sometimes it takes me ten times, but I get there eventually. Thank you everybody for your support on Patreon.

01:17:02
Patreon. com slash AI Inside Show. Thank you to the live.

01:17:05
folks for tuning in also to Amos and Daniel Croft who help behind the scenes for every episode of AI Inside And yeah, I think that's my uh my list of thank yous. Uh Jeff, have a wonderful rest of your week. Everyone uh hang tight and we'll see you next Wednesday on another episode of AI Inside.

01:17:26
Bye. Bye everybody