307: I Hate Smishing

00:00SPEAKERBrad, I posted a podcast from the sky this week.

00:03SPEAKERWhat brought that on?

00:04SPEAKERWell, Linux, I guess, technically.

00:07SPEAKEROkay, I mean, Linux is probably at fault

00:08SPEAKERfor just about everything.

00:10SPEAKERLook, I set the podcast to upload on September 30th

00:14SPEAKERand I scheduled it and I hit the button

00:17SPEAKERand I walked away from the computer

00:18SPEAKERand did some packed up 'cause I was on a trip.

00:21SPEAKERAnd then I didn't check to see that it actually saved

00:25SPEAKERand uploaded and I just closed the laptop.

00:27SPEAKERSo I was like, "Oh, laptop's asleep, must've finished."

00:29SPEAKERAnd then I didn't think about it until the next afternoon

00:32SPEAKERwhen I was on the airplane and I was like,

00:33SPEAKER"Oh, hey, let's see what people think

00:34SPEAKER"about that podcast episode."

00:36SPEAKERI was like, "There's no channel in the Discord."

00:37SPEAKERAnd then I went and looked and I was like,

00:39SPEAKER"Oh, it didn't post."

00:40SPEAKERAnd then I went and looked at the Patreon page,

00:42SPEAKERthe backend CMS, and it was like,

00:45SPEAKER"Yep, no Patreon, no Patreon episode."

00:47SPEAKERSo I uploaded it from the sky.

00:49SPEAKERWell, I guess technically first I downloaded it

00:51SPEAKERfrom my Dropbox and then I uploaded it from the sky.

00:54SPEAKERWow, two-way communication.

00:56SPEAKERIt was not a fast internet connection.

00:58SPEAKERI was getting like 200 kilobits per second.

01:00SPEAKERI bet not.

01:01SPEAKERMy paranoia about everything leads me to triple check

01:04SPEAKERjust about every form before I submit it in general.

01:08SPEAKERUsually I do that, but I was a little distracted

01:10SPEAKERas I was packing and trying to get other stuff done

01:13SPEAKER'cause I was on a trip.

01:15SPEAKERSo yeah, I can't talk about what I saw,

01:16SPEAKERbut I was at the Intel Tech Tour

01:19SPEAKERfor their new processor launches

01:21SPEAKERthat'll be coming at some point in the future.

01:24SPEAKERI'm excited to hear about that.

01:26SPEAKERBut for now, how fast was the download on the airplane?

01:30SPEAKERThe download was okay.

01:32SPEAKERThe download took about eight minutes.

01:33SPEAKERThe upload took pretty much the rest of the two-hour flight.

01:37SPEAKERWe're talking, is this the MP3 I sent you for that podcast?

01:39SPEAKERThis is the, it's a 60 megabyte file.

01:42SPEAKERI got to know each and every bit

01:45SPEAKERas I was downloading and uploading.

01:46SPEAKERSo I was like, "It's a Patreon episode.

01:47SPEAKERHow big can it be?"

01:49SPEAKERIt turns out, 60 megabytes.

01:50SPEAKER61 megabytes, I'm looking at it right now.

01:52SPEAKERYeah, 200 kilobits per second.

01:55SPEAKERDownload?

01:56SPEAKERUpload, it took a minute.

01:58SPEAKERYeah, and by a minute, you mean several minutes?

02:00SPEAKERI mean-

02:01SPEAKERAnd by several minutes,

02:02SPEAKERyou mean like several blocks of several minutes?

02:04SPEAKERWhat I mean is, as they were saying,

02:06SPEAKER"Hey, you need to put your tray table up

02:09SPEAKERso that we can land the airplane."

02:11SPEAKERI was holding the laptop in front of me,

02:13SPEAKERjust barely propped open to get the last 2%.

02:17SPEAKERIt was some real Mission Impossible business, so.

02:19SPEAKERGreat, great.

02:21SPEAKERNo corruption.

02:22SPEAKERAre they using Starlink on those things yet?

02:24SPEAKERSome of them do, but this one did not.

02:26SPEAKERThis was using the old one.

02:28SPEAKERI feel like I've seen that advertised.

02:29SPEAKERNot as in advertising it being in use,

02:32SPEAKERbut like, "Ah, Starlink is coming to the so-and-so,

02:35SPEAKERmaybe Delta fleet of planes," or something.

02:37SPEAKERSo a lot of international stuff has it,

02:40SPEAKERbut this was using the old,

02:41SPEAKERwe point the thing at the cell towers at the ground,

02:44SPEAKERand I believe-

02:46SPEAKERIs that how those things get internet?

02:47SPEAKERThat's how it used to work for a long time.

02:49SPEAKERThe Virgin America internet was,

02:50SPEAKERwe have a big, giant antenna that's bouncing

02:52SPEAKERand sending stuff to LTE towers

02:56SPEAKERon the ground below you. Wow.

02:57SPEAKERI guess that does explain a lot about the speed and latency.

03:00SPEAKERI see all kinds of weird stuff out there.

03:02SPEAKERI don't know how well this works about people, like,

03:04SPEAKERgetting free internet on the plane

03:06SPEAKERby tunneling all of their traffic

03:07SPEAKERthrough whatever port is open that allows you to,

03:10SPEAKERyou know, you have to do their captive portal

03:12SPEAKERand pay for service,

03:15SPEAKERpeople finding ways to tunnel all of their traffic

03:17SPEAKERthrough whatever openings there are that allow that stuff.

03:20SPEAKERMy Linux computer happily connected

03:23SPEAKERto the broad internet without paying.

03:25SPEAKERAh, really?

03:26SPEAKERUh-huh. Interesting.

03:28SPEAKERI think they're just getting DNS, maybe.

03:31SPEAKERI'm not saying I would necessarily attempt

03:32SPEAKERsomething like that, but-

03:33SPEAKERNo, I paid the $8. It was worth it.

03:35SPEAKERIt's interesting that that sounds

03:37SPEAKERlike it might be possible, anyway.

03:38SPEAKERThis podcast does not endorse any hijacking

03:41SPEAKERof internet service through dot, dot, dot.

03:45SPEAKERThat's a federal crime, probably,

03:46SPEAKER'cause it's on an airplane.

03:47SPEAKERI'm not gonna fool around with that.

03:48SPEAKERI'll pay the $8.

03:49SPEAKERLet's just leave that alone, just pay the $8.

04:20SPEAKERWelcome to "Brad and Will Made a Tech Pod."

04:22SPEAKERI'm Will.

04:23SPEAKERI'm Brad.

04:24SPEAKERHello. Hi, hello.

04:25SPEAKERHello. Welcome back into town.

04:27SPEAKERI am excited to be back.

04:28SPEAKERIt has been a whirlwind week.

04:30SPEAKERI saw a roadrunner, a coyote, a scorpion.

04:34SPEAKERI just started to ask if you can say

04:36SPEAKERwhat state this was in, but I don't think I have to ask now.

04:38SPEAKERWhere do you think I was?

04:39SPEAKERI'm guessing Arizona.

04:41SPEAKERYeah, that's it. That's it.

04:42SPEAKERI saw a hawk.

04:43SPEAKERWe got hawks right out back of here,

04:45SPEAKERbut scorpions, that's a roadrunner?

04:46SPEAKERThat's a different thing.

04:47SPEAKERYeah, the roadrunner's really truly hideously ugly animals.

04:51SPEAKEROf all the animals that got a cartoon glow up,

04:53SPEAKERthe roadrunner is the king.

04:54SPEAKERReally?

04:55SPEAKERIs an absolutely revolting looking bird.

04:57SPEAKERWhat?

04:58SPEAKERYeah, and every time I've ever seen a roadrunner in person,

05:01SPEAKERthey always have some dead lizard

05:03SPEAKERdangling out of their mouth.

05:04SPEAKEROkay, I started to say, what's revolting about this bird?

05:07SPEAKERIt looks kind of noble,

05:08SPEAKERbut then I saw a closeup of its face.

05:11SPEAKERThey're not attractive animals.

05:12SPEAKERNow I see what you're saying.

05:14SPEAKERThat's a face only a mother could love.

05:15SPEAKERYeah, don't come at me, Arizonans.

05:19SPEAKERIt's fine, whatever.

05:20SPEAKERDid you see it running?

05:22SPEAKERNo, it was just kind of walking.

05:23SPEAKEROh.

05:24SPEAKERI know.

05:27SPEAKERRight now, it's September.

05:29SPEAKERI was there in September.

05:30SPEAKERIt was so hot during the day that the evenings were even hot.

05:34SPEAKERSo the only time you'd go outside

05:36SPEAKERwas seven o'clock in the morning.

05:38SPEAKERSo I went out for a walk a couple of mornings

05:39SPEAKERwhile I was there when it was still 70 degrees

05:41SPEAKERand really nice.

05:42SPEAKERAnd I saw the roadrunner and the coyote

05:44SPEAKERon two separate days.

05:46SPEAKERThe coyotes there are bigger than our coyotes here

05:48SPEAKERis the first thing I noticed.

05:50SPEAKEROur coyotes here are pretty small.

05:51SPEAKERYou see a lot of coyotes around here to judge?

05:54SPEAKERDude, I've seen so many coyotes

05:56SPEAKERin the last two years, yeah.

05:57SPEAKERWow.

05:58SPEAKERWe saw one, when we were camping in the park

06:01SPEAKERa couple of weeks in the Presidio a few weeks ago,

06:03SPEAKERwe saw one.

06:04SPEAKERWow.

06:05SPEAKERI see them down in Pacifica kind of regularly.

06:07SPEAKERIf you're up in the mornings or out,

06:09SPEAKERlike duskish is when they seem to be out the most.

06:11SPEAKERAnd I hear them all the time.

06:12SPEAKERYou only ever see single ones?

06:14SPEAKERIt depends on the time of the year.

06:15SPEAKERSo you see single ones apparently when the cubs,

06:19SPEAKERthat year's cubs are kicked out of the pack.

06:21SPEAKERWhen they reach adolescence, the males, they kick them out

06:24SPEAKERand then they have to go find their own people

06:26SPEAKERor I guess coyotes.

06:28SPEAKERI've always been under the impression

06:30SPEAKERthat coyotes are worst in packs.

06:33SPEAKERPacks are when they get really dangerous

06:35SPEAKERis my understanding.

06:36SPEAKERYeah, you don't, I don't, we hear packs,

06:39SPEAKERwe hear them talking to each other across the valley

06:40SPEAKERin Pacifica fairly regularly,

06:42SPEAKERbut I've never seen more than one, one or two together.

06:45SPEAKERSo.

06:46SPEAKEROkay, like I've always heard like,

06:47SPEAKERthat's how they will corner and kill dogs, for example,

06:49SPEAKERis when like one, one coyote will drive the dog

06:53SPEAKERback into the pack and then that's a bad scene.

06:56SPEAKERYeah, I don't, I don't, I've seen,

06:58SPEAKERlike there are a lot of signs up in the Presidio

07:00SPEAKERand in Golden Gate Park now that are,

07:01SPEAKERthat are like, hey beware, cause there's coyotes here.

07:05SPEAKERAnd we saw them, I've seen them like walking around

07:07SPEAKERdown by Bayshore, like in the hills up above Bayshore.

07:12SPEAKERI've seen them walking around.

07:14SPEAKERI saw one down by Chase Center one night actually,

07:17SPEAKERjust like casually crossing the street

07:19SPEAKERdown by the, down by where the Warriors

07:22SPEAKERand the Valkyries play, so.

07:24SPEAKERThink he was a Warriors fan.

07:25SPEAKERI assume he was a Valkyries fan, it was after the game, so.

07:27SPEAKERYeah, sure.

07:29SPEAKERAnyway, we got a news roundup this week.

07:31SPEAKERWe got three hot stories.

07:34SPEAKERI think these are three,

07:35SPEAKERthree things kind of struck our interest.

07:38SPEAKERYou wanna start?

07:39SPEAKERWell, I guess, should we tell people what they are?

07:41SPEAKERSure, yeah.

07:42SPEAKERLet's provide a menu up top here that they can then,

07:45SPEAKERI guess, like randomly seek through their podcast to find.

07:48SPEAKERYeah, the first one is about smishing,

07:50SPEAKERwhich I thought was like when you smile and wish

07:53SPEAKERat the same time, maybe.

07:55SPEAKERYes.

07:56SPEAKERBut it's SMS phishing, which is a new term for me.

08:00SPEAKERPhishing already, some kind of portmanteau or mashup of,

08:04SPEAKERwhy are cyber crime terms so goofy, I guess,

08:07SPEAKERis the real question.

08:08SPEAKERI think it's cause the kind of people who cover

08:10SPEAKERthat kind of stuff are goofballs.

08:12SPEAKERYeah, that's very possible.

08:14SPEAKERThat's kind of how we got bootstrapping

08:16SPEAKERor booting a computer, right?

08:18SPEAKERReally?

08:19SPEAKEROh yeah, yeah.

08:19SPEAKERThat comes from, I don't know how to pronounce it,

08:21SPEAKERBaron, Baron, help me out, Munchausen.

08:24SPEAKERMunchausen?

08:25SPEAKERMunchausen, I don't know how people say that,

08:27SPEAKERbut I believe that's where that comes from.

08:28SPEAKERIt comes from pulling yourself up by your bootstraps.

08:30SPEAKERSame place as the other use of that term.

08:34SPEAKERYeah, the thing we all do.

08:35SPEAKEROr, you know, there's like so many terms

08:36SPEAKERthat come out of Douglas Adams or Monty Python.

08:39SPEAKEROh yeah, nerds.

08:40SPEAKERThere is just, yes.

08:41SPEAKERThere is a specific kind of like '70s, '80s nerd humor

08:45SPEAKERthat has informed so much of computing terminology.

08:47SPEAKERAnyway, two different stories about the phenomenon of SMS,

08:52SPEAKERlike text phishing and spam and related cyber crime

08:55SPEAKERhave both come out around the same time

08:57SPEAKERthat kind of caught our interest.

08:58SPEAKERYeah, so then the next story is Ars Technica flagged

09:01SPEAKERthis story about the iRobot/Rumba founder

09:06SPEAKERwho's been doing robotics since the '70s

09:09SPEAKERwho won't go within nine feet of today's walking robots.

09:13SPEAKERWhich, like that was interesting enough,

09:15SPEAKERbut then the blog post from Rodney Brooks

09:18SPEAKERthat they linked to that's the source for that

09:20SPEAKERis fascinating, we'll talk about that.

09:23SPEAKERAnd then the last thing is we're like two,

09:26SPEAKERthis is, it's not this podcast,

09:27SPEAKERbut the next podcast I think is the last one

09:29SPEAKERbefore Windows 10 goes end of life.

09:31SPEAKERYes, correct.

09:34SPEAKERAnd there's some downstream consequences from that

09:35SPEAKERthat I hadn't really thought about, but it makes sense.

09:38SPEAKERYeah, some software getting retired in advance

09:41SPEAKERor support ending for some software

09:44SPEAKERthat depends on old versions of Windows, let's say.

09:46SPEAKERLet's say it marks the end of an era.

09:48SPEAKERYeah, yes, say that.

09:51SPEAKERSo you wanna start with smishing?

09:54SPEAKERYes, I think there's a lot of meat on that bone.

09:56SPEAKERCan I just say I hate smishing,

09:59SPEAKERlike the name and everything about it.

10:01SPEAKERThe name's bad.

10:02SPEAKERYeah, I don't think it's just the name.

10:03SPEAKERI don't think there's anything that anybody likes about this.

10:05SPEAKERMaybe somebody does, I don't know.

10:07SPEAKERMaybe occasionally you give one of these fake text messages

10:09SPEAKERfrom somebody going like, hey, are you free right now?

10:11SPEAKERAnd you go like, oh man, somebody cares about what I'm doing.

10:14SPEAKERThey wanna hang out.

10:15SPEAKERHi, I thought I'd reach out and say hi

10:17SPEAKERor say hi to Mike for me.

10:19SPEAKERLike what, you're boarding your flight,

10:22SPEAKERyou're now boarding your flight to San Francisco at gate E7.

10:24SPEAKERThat's not a real thing.

10:25SPEAKEROh, wait, no, I actually wanted, that one was real.

10:27SPEAKERI'm pretty sure that one was actionable.

10:28SPEAKERThat one was right.

10:29SPEAKERAnd intended recipient and so on.

10:31SPEAKERI got an amazingly specific one yesterday.

10:34SPEAKEROh, okay.

10:35SPEAKERIt's been a while since we last met

10:37SPEAKERand I miss you so much.

10:38SPEAKERI wonder how you're doing lately.

10:40SPEAKERThe weather is beautiful this weekend

10:41SPEAKERand I wonder if I could have you on the course

10:43SPEAKERfor a round of golf.

10:45SPEAKERWow.

10:46SPEAKERThat's funny.

10:47SPEAKERThat's the random smish I got yesterday.

10:49SPEAKERI got one that said fogs out, hogs out.

10:52SPEAKERSo I was like, no, that would be amazing though.

10:56SPEAKERThat would be pretty good.

10:57SPEAKERWhy the golf part?

10:59SPEAKER'Cause people play golf.

11:00SPEAKERYeah, but like are they just trying to connect with you

11:03SPEAKERon a subject that you might be passionate enough about

11:05SPEAKERto respond?

11:06SPEAKERI assume.

11:08SPEAKERSo, okay.

11:09SPEAKERSo I've talked to some security experts about this.

11:12SPEAKERWe had our security guy at PC World talked about this.

11:16SPEAKERLike the messages are so random.

11:19SPEAKERMost of them, when they hook you,

11:21SPEAKERthey feed you into like big butchering schemes

11:23SPEAKERwhere people start trying to get,

11:25SPEAKERfirst they go for engagement

11:26SPEAKERand then they just keep working, grinding you down

11:28SPEAKERuntil suddenly you're giving them thousands upon thousands

11:31SPEAKERof dollars and you don't exactly understand how it happened.

11:34SPEAKERAnd the common thread in coverage of those has been

11:38SPEAKERthat even security experts are surprised at how easy it is

11:41SPEAKERto happen to them.

11:43SPEAKEREven people who know how this works,

11:45SPEAKERhow the psychology of it works, how it happens,

11:47SPEAKERit's still easy to mess up and become engaged with this.

11:52SPEAKERNow, I don't know if I believe that there's enough people

11:56SPEAKERthat will buy into this stuff that it actually,

12:01SPEAKERthat it's actually a profitable enterprise.

12:03SPEAKERBut then you read stories about things like people

12:05SPEAKERwho have Bitcoin ATMs in their convenience store,

12:07SPEAKERhaving to shut them off at certain times of day

12:09SPEAKERbecause elderly people are coming in desperate

12:12SPEAKERto feed money into the machines and stuff like that

12:14SPEAKERbecause of these kinds of scams.

12:16SPEAKER- It would have to be profitable on an ongoing basis

12:19SPEAKERfor them to keep doing it, right?

12:20SPEAKERWell, I mean, as we'll see, as we get into this,

12:22SPEAKERthese methods of massively scaling the capabilities

12:26SPEAKERof SIM cards and cellular connections have a lot

12:28SPEAKERof other uses besides garden variety scamming,

12:31SPEAKERbut still like they have to be making enough money on this

12:34SPEAKERto make it worth it.

12:35SPEAKER- Honestly, the thing that annoys me about this more

12:38SPEAKERthan anything else is that I used to be excited

12:42SPEAKERwhen I got a text message, remember that?

12:45SPEAKER- Sure, sure, yes, yes, okay, yes.

12:48SPEAKER- Remember you were like-

12:49SPEAKER- Yes, if we get back far enough and I also happened

12:51SPEAKERto be in my 20s then and had a social life

12:53SPEAKERand like to go out and do things a lot and stuff, sure.

12:57SPEAKER- I made the text message beep, the Galaga insert coin sound

13:01SPEAKERbecause I was like, it's fun when I get a text message.

13:03SPEAKERI want it to sound fun and I want my brain to think,

13:05SPEAKERI'm about to have some fun.

13:07SPEAKERAnd now-

13:08SPEAKER- I made it the Metal Gear, you got caught sound.

13:11SPEAKER- Yeah, yeah, another popular one

13:12SPEAKERin the Whiskey Media office in like 2009 to 2010.

13:15SPEAKER- Yes, the heyday of getting text messages

13:17SPEAKERand enjoying it.

13:18SPEAKER- Yeah, hey, we're at the bar across the street,

13:20SPEAKERyou wanna come over?

13:21SPEAKERYes, I would love to go to the bar across the street

13:23SPEAKERand hang out for a little bit.

13:25SPEAKERNow it's just, I get a text message and it's a sense of dread

13:27SPEAKER'cause it's either Verizon telling me you're gonna hit me

13:29SPEAKERwith a bill or somebody trying to scam me

13:33SPEAKERfor something I don't have probably.

13:34SPEAKERAnyway, this came up because of Wired story.

13:38SPEAKERAll these stories are linked in the show notes.

13:41SPEAKERThe Secret Service busted a SIM farm

13:46SPEAKERthat was a kind of insane scale.

13:49SPEAKERLike 100,000 SIM cards in these servers.

13:53SPEAKERAnd like when I read the headline for this story,

13:55SPEAKERI was expecting, in my head, I was seeing pictures

13:57SPEAKERof those old school Southeast Asian spam farms

14:01SPEAKERwhere you have like one person sitting in front of a rack

14:03SPEAKERwith 150 phones on it and they're just tapping each phone

14:07SPEAKERevery 30 seconds, right?

14:08SPEAKER- Or in some cases, yeah, or in some cases,

14:10SPEAKERsome of those that I've seen going around are not even,

14:12SPEAKERthey're not even fully assembled mounted phones

14:16SPEAKERwith like screens that can be interacted with.

14:18SPEAKERThey've been like, they're like gutted Android phones

14:20SPEAKERthat are just like the innards of the phones

14:23SPEAKERare slammed into racks and are being used for some,

14:26SPEAKERwhat is that, scale?

14:28SPEAKER- And often it was like, it looked,

14:31SPEAKERI don't wanna say slapdash 'cause that's not exactly right,

14:33SPEAKERbut they were like PVC pipes

14:36SPEAKERwith a bunch of phones zip tied to them.

14:38SPEAKERSo it wasn't, it didn't look like,

14:40SPEAKERit didn't look like what I would think of

14:41SPEAKERas an organized rack.

14:42SPEAKERIt looked like something that somebody hacked together

14:45SPEAKERand plugged a whole bunch of power into

14:46SPEAKERand then somebody's hellish job

14:49SPEAKERbecomes to click on these things.

14:50SPEAKEROften worth mentioning those old spam farms

14:53SPEAKERare involved with human trafficking as well.

14:55SPEAKERSo this was different.

14:57SPEAKERThe racks for these things looked like,

14:59SPEAKERthey almost looked like those,

15:01SPEAKERthey looked almost like an audio mixer or something.

15:03SPEAKERIt was like a row after row after row of SIM cards

15:07SPEAKERwith a bunch of antennas around the edges.

15:09SPEAKER- Yeah, you're right.

15:10SPEAKERLooking again at the photo, I see what you mean.

15:12SPEAKERLike the console is kind of like

15:14SPEAKERa giant mixing board almost.

15:16SPEAKER- Yeah, or it looks like,

15:17SPEAKERit looks like if you need to duplicate data

15:21SPEAKERacross a whole bunch of USB thumb drives,

15:23SPEAKERyou buy this thing that has like 200 ports

15:25SPEAKERand then you just jam ports

15:26SPEAKERand then you copy them to all of them at once in parallel.

15:28SPEAKER- Yeah, but the key thing here though is that these,

15:30SPEAKERI think SIM box is the term they use colloquially in this,

15:34SPEAKERI don't know what the actual name of the equipment is,

15:36SPEAKERbut that's what they refer to them as in this wired story.

15:39SPEAKERBut those are illegal.

15:40SPEAKERLike SIM boxes are not legal in America, so.

15:43SPEAKER- Weird.

15:43SPEAKER- Yeah, but it seems like these have been smuggled in.

15:46SPEAKERThis was, I don't know if you mentioned,

15:47SPEAKERthis was all in the tri-state New York City area.

15:50SPEAKER- Yeah, all they said was that

15:51SPEAKERit was within 35 miles of Manhattan.

15:54SPEAKERAnd then they said a bunch of really exciting things like,

15:57SPEAKERoh, this was involved in the swatting attacks

16:00SPEAKERon Congress members in December of 2003 over Christmas.

16:03SPEAKER- Or 2023, yeah.

16:05SPEAKER- And the New York Secret Service said

16:06SPEAKERthat the scale of this installation

16:09SPEAKERwould have let them overwhelm cell towers

16:11SPEAKERand essentially shut down the cellular network

16:13SPEAKERin New York City 'cause they were able to send

16:15SPEAKER30 million text messages per minute

16:17SPEAKERwith the hardware that they had.

16:19SPEAKER- Yeah, this quote from the Secret Service agent says,

16:22SPEAKERto put it in context, this could have been used

16:24SPEAKERto text every number in the United States

16:26SPEAKERin about 12 minutes.

16:28SPEAKER- Yeah, so who uses this stuff, Brad, do we know?

16:32SPEAKER- As you said, I think they got on the trail for this

16:34SPEAKERbecause of the swatting attacks against Congress people,

16:37SPEAKERwhich I don't remember, was that?

16:38SPEAKER- I don't know that they made a big deal about it.

16:40SPEAKERIt was Marjorie Taylor Greene and Jim Jordan, I think.

16:43SPEAKERRick Scott, I think is the other one that they mentioned

16:45SPEAKERhere, anyway, that sounds like that's what got

16:47SPEAKERthe Secret Service on the trail of this.

16:49SPEAKERObviously it's used for mass amounts of the random

16:54SPEAKERscamming, phishing type text messages

16:57SPEAKERthat we have talked about on here already.

17:01SPEAKERSwatting is another one, like this is,

17:03SPEAKERI mean, obviously, swatting against Congress people,

17:06SPEAKERwhat I mean is like swatting generally.

17:07SPEAKERApparently this is used as kind of a swat for hire middleman

17:12SPEAKERfrom the sound of things, or it's used as like

17:14SPEAKERinfrastructure for like professional.

17:18SPEAKERBut again, I know we already said the Congress people

17:20SPEAKERthing feels like its own separate thing,

17:22SPEAKERbut it sounds like this could also be used

17:23SPEAKERto swat kind of anybody at will.

17:26SPEAKER- Well, the implication is that this probably

17:29SPEAKERwouldn't have been tracked down had it not been used

17:31SPEAKERin the Congress people thing.

17:32SPEAKERThat was the thing that gave them the incentive

17:33SPEAKERto go after it 'cause they were worried

17:35SPEAKERthat it might be a nation state trying to interfere

17:37SPEAKERwith the UN meeting that happened last week

17:39SPEAKERor the week before.

17:40SPEAKER- Yes, right, yes.

17:41SPEAKERThat the amazing thing to me is that this thing

17:44SPEAKERcan send out, you know, text the entire population

17:48SPEAKERof the United States in 12 minutes,

17:50SPEAKERand that it took swatting some Congress people

17:53SPEAKERto make it worth tracking down.

17:55SPEAKER'Cause it's radio, it's not like they don't know

17:58SPEAKERwhich cells these boxes are connecting to, right?

18:01SPEAKERYou'd think even that the network could say,

18:05SPEAKERoh yeah, all of a sudden we're getting 3,000 times

18:07SPEAKERor 3 million times the traffic per cell in this region.

18:12SPEAKERMaybe we should call the FBI or the FCC

18:15SPEAKERor somebody and track it down.

18:17SPEAKER- Yeah, I think one of the things they mentioned

18:18SPEAKERin the story is these systems are very good

18:20SPEAKERat rotating SIM cards and phone numbers in and out,

18:24SPEAKERpresumably to deal with maybe rate limiting

18:26SPEAKERand tracking and that kind of thing.

18:28SPEAKER- The law enforcement officers, they quoted,

18:30SPEAKERsaid it basically, like predominantly this is used

18:33SPEAKERfor fraud and scams, and then they do also sell services

18:37SPEAKERout to other folks as well.

18:40SPEAKER- Yeah, yeah, I found the quote right here.

18:41SPEAKERThe technology behind these farms

18:43SPEAKERmakes them highly flexible.

18:44SPEAKERSIMs can be rotated to bypass detection systems,

18:47SPEAKERtraffic can be geographically masked,

18:49SPEAKERand accounts can be made to look like they're coming

18:51SPEAKERfrom genuine users.

18:52SPEAKERSo these things are very good

18:54SPEAKERat covering their tracks apparently.

18:56SPEAKER- I don't understand how they can geographically mask

18:58SPEAKERtheir connection when they're clearly,

19:01SPEAKERthey have to connect to the cell

19:02SPEAKERthat's in the region where they are.

19:04SPEAKER- Maybe that means like IP masking, like.

19:07SPEAKER- Maybe.

19:08SPEAKER- Maybe defeating IP geolocation

19:10SPEAKERand not like physically tracking the source of a signal,

19:13SPEAKERI would assume.

19:14SPEAKER- So they said each SIM box had 256 ports

19:17SPEAKERand associated modems for each of those cards.

19:20SPEAKERAnd they showed stacks and stacks and stacks

19:22SPEAKERof punched out SIM cards from Mobile X,

19:25SPEAKERwhich is a burner SIM card company.

19:29SPEAKERIt makes me wonder some stuff about this.

19:31SPEAKERLike I wonder, I don't know if you've seen lately,

19:33SPEAKERbut they recently started adding the verified number thing

19:37SPEAKERto your phone calls and text messages

19:40SPEAKERwhen you get them on iOS.

19:42SPEAKERSo it'll say, like if you look at your call log history

19:45SPEAKERor your text, I guess call log history is the main one,

19:48SPEAKERbut it'll say, it's not opening, there we go.

19:54SPEAKERIt'll say potential spam, or when you click on one

19:57SPEAKERthat's not potential spam, it'll show you

20:01SPEAKERa little check mark that means,

20:04SPEAKERoh hey, this person connected to the network

20:08SPEAKERin a meaningful, calls with a check mark

20:11SPEAKERhave been verified by the carrier.

20:12SPEAKERSo I presume that that's the difference

20:14SPEAKERbetween burner people, burner calls,

20:16SPEAKERand like a Verizon or AT&T contract number call.

20:21SPEAKER- Is that new in iOS 26?

20:23SPEAKER- No, it came in at least last year, maybe the year before.

20:26SPEAKER- Okay.

20:27SPEAKER- It's in the contact, if you go to the contact setting

20:31SPEAKERfor the call, for a call that came in in your recents.

20:35SPEAKERBut I have a ton of spam calls with those check marks,

20:37SPEAKERso clearly something's not working.

20:40SPEAKERI also wondered if eSIM helps with this,

20:42SPEAKERand I dug into that a little bit,

20:44SPEAKER'cause presumably if you can't go buy a card

20:48SPEAKERthat you then jam into this thing,

20:49SPEAKERthen you'd have to have eSIMs,

20:50SPEAKERand those eSIMs have IMEIs attached to them

20:52SPEAKERand all that stuff, so that they're permanently attached

20:55SPEAKERand if they get banned by the network

20:57SPEAKERthen they're banned forever.

20:58SPEAKERBut it seems like, yeah maybe, but also realistically,

21:03SPEAKERthey're gonna be selling, there's enough phones

21:05SPEAKERon the market that require SIM cards

21:06SPEAKERthat we're gonna be selling SIM cards in real life

21:08SPEAKERfor the rest of our meaningful lives, it seems like.

21:11SPEAKER- So, the problem with eSIM is that they want it

21:16SPEAKERto be as usable as a normal plastic SIM card,

21:21SPEAKERso you still can, like when I traveled to Canada

21:24SPEAKERI was able to buy an eSIM card for like $25,

21:26SPEAKERand again, an eSIM that connected to my phone

21:29SPEAKERhad like 50 bucks worth of cell coverage or whatever in it

21:32SPEAKERthat was good for the entire time I was there, so.

21:34SPEAKER- So, now I've only bought a phone once

21:36SPEAKERthat had an eSIM on it,

21:37SPEAKERso I've only been through this process once,

21:39SPEAKERbut you have to go through some kind of carrier

21:40SPEAKERor infrastructure to access an eSIM, right?

21:45SPEAKER- You have to, yeah, whoever you buy your eSIM from,

21:49SPEAKERyou have to authorize it and connect it to the modem

21:51SPEAKERand stuff on the device, yeah.

21:52SPEAKER- Right, and you couldn't just move it from one phone

21:55SPEAKERto another on your own without going through

21:57SPEAKERtheir infrastructure, right?

21:58SPEAKEROr again, maybe there is some way now.

22:00SPEAKER- Yeah, but you also, having done the moving your eSIM card

22:03SPEAKERfrom one phone to another fairly regularly

22:05SPEAKERin the tested days, you usually have to authorize it

22:09SPEAKERon the, even when you switch the physical card,

22:12SPEAKERyou have to go in and say, yes, I did this,

22:14SPEAKERthis is a new device, please make this my phone,

22:19SPEAKERthe thing that has the card in it that says it's my phone.

22:21SPEAKER- Okay. - It's not as easy

22:22SPEAKERas the old 3G days where you just swapped it over

22:25SPEAKERand it didn't know. - Oh, interesting.

22:26SPEAKEROkay, I have not swapped a SIM card since then,

22:28SPEAKERthat is good to know.

22:29SPEAKER- So part of it is that the SIM card hacking/cloning

22:38SPEAKERis one of the big fraud things,

22:40SPEAKERso they've tightened up how that works fairly significantly.

22:43SPEAKERPart of that may be also 'cause I have my Verizon,

22:45SPEAKERmy phone account locked down, so.

22:48SPEAKER- Sure.

22:48SPEAKER- Anyway, there was another follow-up on this,

22:51SPEAKERanother shout out to Ars Technica

22:53SPEAKERfor really blowing it out of the water this week.

22:56SPEAKERThey had a story about unsecured internet of things devices

23:00SPEAKERthat have cellular routers and cellular modems in them.

23:03SPEAKER- Yeah, this one somehow is more mind blowing to me

23:06SPEAKERthan the organized crime, like highly sophisticated

23:10SPEAKERregimented operation doing, telecom cyber crime, of course,

23:15SPEAKERvulnerable devices sitting around hooked up

23:17SPEAKERto traffic lights and infrastructure

23:19SPEAKERthat people are using to do this stuff,

23:21SPEAKERlike that one is, that one feels more insidious to me

23:24SPEAKERbecause who knows how many vulnerable devices are out there

23:27SPEAKERor how extensively this stuff could be exploited.

23:29SPEAKER- Look, I never thought that I would agree

23:31SPEAKERwith the people in Marin that are like,

23:33SPEAKERkeep your smart meters out of my yard, man,

23:35SPEAKERthey're killing my brain, they got the 5G.

23:38SPEAKER- Well, I started to say maybe they had a point,

23:40SPEAKERbut I'm gonna say that's probably not the point

23:41SPEAKERthey thought they had.

23:42SPEAKER- Yeah, they had the, I mean, sometimes the, you know,

23:45SPEAKERthe worst person you know is right sometimes.

23:48SPEAKERYeah, so that's the situation.

23:50SPEAKERThere were a bunch of these devices out in the world

23:52SPEAKERwith a ton of known vulnerabilities that don't get updated.

23:55SPEAKERIt's in stuff like smart meters, traffic lights,

23:57SPEAKERlike you said, monitoring infrastructure for utilities

24:02SPEAKERand traffic and all sorts of other stuff, industrial stuff,

24:05SPEAKERthat's connected to cellular because it's away

24:08SPEAKERfrom a dedicated network connection

24:09SPEAKERbecause it has a backup cellular connection.

24:12SPEAKERAnd these things are responsible for apparently billions

24:15SPEAKERof messages a month in scam SMS messages.

24:19SPEAKER- Right, so this came out of this security research firm

24:22SPEAKERcalled Sequoia, and it's focused on this one model

24:24SPEAKERof this one company's device.

24:26SPEAKER- Great.

24:27SPEAKER- This is a UR35 cellular router made by a company

24:31SPEAKERcalled Milesight, and they go into, you know,

24:33SPEAKERthis article goes into like, oh, specific firmware versions

24:35SPEAKERearlier than this are rife with vulnerabilities.

24:38SPEAKERAnd of course, all these devices that are being exploited

24:40SPEAKERare on old firmware that is not being updated.

24:43SPEAKERThat's kind of what I meant about this feeling

24:44SPEAKERmore insidious though, is because again, this is like,

24:46SPEAKERthis is about one manufacturer's product line.

24:49SPEAKERWho knows how many other competing manufacturers

24:51SPEAKERthere are out there with similar vulnerabilities

24:53SPEAKERin devices that, and again, this is the critical failure,

24:56SPEAKERare not being maintained and not being updated.

24:58SPEAKER- I mean, the thing that sucks about both of these

25:00SPEAKERis that they're not tractable.

25:02SPEAKERLike there's no easy fix for either of these.

25:05SPEAKERYou know, the hardware to put 100,000 SIM cards on a node

25:10SPEAKERseems inexpensive enough that it could just be abandoned

25:13SPEAKERin a fake data center, in a malware data center.

25:17SPEAKER- Not even that, I mean, I think the last line

25:19SPEAKERof this R story kind of summed it up nicely

25:23SPEAKERwhen they're talking about,

25:23SPEAKERwell, I'll just read the whole paragraph.

25:25SPEAKERGiven the prevalence and massive volume

25:26SPEAKERof smishing messages, people often wonder how scammers

25:29SPEAKERmanage to send billions of messages per month

25:31SPEAKERwithout getting caught or shut down.

25:33SPEAKERSequoia's investigation suggests that in many cases,

25:35SPEAKERthe resources come from small, often overlooked boxes

25:38SPEAKERtucked away in janitorial closets in industrial settings.

25:43SPEAKER- Right.

25:43SPEAKER- Like think about how many cases there must be

25:45SPEAKERwhere somebody, like some overworked maintenance

25:48SPEAKERor infrastructure person goes out

25:49SPEAKERand has to deploy some equipment,

25:51SPEAKERtacks it to the wall in whatever place they can find,

25:54SPEAKERlike runs a cable through a gap in the wall

25:57SPEAKERand then just forgets it.

25:58SPEAKER- Yeah, never thinks about it again.

26:00SPEAKER- Yes, everybody associated just forgets that it's there.

26:02SPEAKER- Fantastic.

26:04SPEAKER- So that's just the idea of infrastructure

26:06SPEAKERjust sitting around and being exploited by bad actors is,

26:10SPEAKERit's kind of cool and exciting in one way,

26:12SPEAKERexcept for the part where they're using it

26:14SPEAKERfor all kinds of bad shit, you know?

26:16SPEAKER- Yeah, I mean, the fact that malware is scaling faster

26:19SPEAKERthan we can improve our ability to deal with it is an issue.

26:24SPEAKERLike, I think I've told this story here before,

26:26SPEAKERbut there was a day, I think it was after

26:29SPEAKERthe Whiskey Media sale when they had used

26:33SPEAKERan insecure comment script for the new tested website.

26:36SPEAKERIt was a WordPress or Drupal plugin or something.

26:40SPEAKERAnd we just started getting thousands of spam messages

26:45SPEAKERposted on every blog post on the site,

26:48SPEAKERthousands upon thousands of blog posts every minute.

26:52SPEAKERAnd it was while Norm and I were at Maker Faire in New York

26:54SPEAKERand there was 30 people deleting them

26:58SPEAKERas fast as they could come in

26:59SPEAKERbefore we finally had the DevOps guy figure out

27:02SPEAKERhow to turn off comments for the day, right?

27:05SPEAKERAnd there's just no, it's impossible to deal with that scale

27:10SPEAKERif people are able to make money off of those messages.

27:13SPEAKERSo I don't know how we fix it.

27:15SPEAKER- This reminds me of a story.

27:17SPEAKERI just pulled it up as we've been recording.

27:18SPEAKERIt's been, this is about a year and a half old

27:20SPEAKERand it's been that long since I read it.

27:21SPEAKERSo I apologize, I don't have all the details front of mind.

27:24SPEAKERYou may have seen this, this is actually on ours as well.

27:26SPEAKERThe DOJ, Department of Justice about a year and a half ago

27:30SPEAKERcame out and talked about compromised ubiquity routers.

27:34SPEAKER- Oh, I do remember this.

27:35SPEAKER- More than a thousand ubiquity routers in homes

27:38SPEAKERand small businesses were infected with malware

27:40SPEAKERused by Russian backed agents to coordinate them

27:43SPEAKERinto a botnet for crime and spy operations,

27:45SPEAKERaccording to the justice department.

27:47SPEAKERAnd like whatever hacker group had gotten in

27:49SPEAKERand compromised these things through vulnerability.

27:51SPEAKERI think the part that really,

27:53SPEAKERpart that I really laughed at is that supposedly

27:55SPEAKERthe US government was then going out

27:56SPEAKERand using the same vulnerabilities

27:58SPEAKERto silently go into these same individual

28:01SPEAKERand small businesses routers.

28:02SPEAKER- To fix them, right? - Yes, to update them

28:04SPEAKERto kind of shut off the vulnerability and update them

28:08SPEAKERand prevent this from happening.

28:10SPEAKERLike it's just this like cat and mouse game.

28:12SPEAKER- I mean, maybe that's the solution.

28:13SPEAKERIt's interesting 'cause like the router saves people,

28:17SPEAKERhaving a router that updates itself

28:19SPEAKERand you don't have to think about saves a lot of people

28:22SPEAKERwho don't pay attention to this stuff

28:24SPEAKERfrom an enormous number of problems, right?

28:26SPEAKERLike if you think, it's one of the reasons

28:28SPEAKERthat I actually kind of like having AT&T provide my router.

28:31SPEAKERSure, it's the same hardware every single other person

28:33SPEAKERon the network has and sure I could put something else there

28:35SPEAKERbut it'd be kind of a pain in the butt.

28:37SPEAKERBut also every 30 days it updates itself

28:41SPEAKERand I don't have to worry about,

28:42SPEAKERlike I know that I'm as protected as the might

28:46SPEAKERof AT&T's fiber internet security ops people can make me.

28:52SPEAKERAnd if I were running a PF sense or open sense

28:54SPEAKERor something like that, that would be on me.

28:56SPEAKERSo while I'm capable of doing that,

28:59SPEAKERit's nice to not have to.

29:00SPEAKERAnd I think when we talk about the Windows 10

29:02SPEAKERend of life stuff, the widespread existence of routers

29:06SPEAKERthat protect people from the raw internet

29:08SPEAKERis maybe gonna be the thing that saves Microsoft

29:10SPEAKERbut we'll talk about that in a little bit.

29:11SPEAKER- To be clear, the same thing is true.

29:13SPEAKERIf you have a consumer router as well,

29:14SPEAKERyou still need to keep it updated.

29:15SPEAKERAlthough I don't know, do consumer routers ship

29:17SPEAKERwith auto updates as an option now?

29:21SPEAKER- So the mesh stuff that I've tested

29:22SPEAKERand help people set up generally does.

29:24SPEAKERIt works like your phone does where it knows

29:27SPEAKERwhen your downtime is and at two o'clock in the morning

29:29SPEAKERit'll run an update when there's one available.

29:32SPEAKERI don't know about the kind of in-between ones

29:34SPEAKERlike your ASUS or Netgear router with 12 antennas.

29:38SPEAKERI assume that they do.

29:39SPEAKER- Or even lower tier than that.

29:41SPEAKERI feel like your $90 TP-Link Archer

29:44SPEAKERis like the common cheap router recommendation these days.

29:47SPEAKERI don't know if those auto update or not.

29:49SPEAKER- The closer you get to the cheap end,

29:51SPEAKERthe more likely they are to update

29:53SPEAKER'cause the less likely they are to want

29:55SPEAKERto have to have secure calls, support calls.

29:58SPEAKER- Yeah, that would make sense.

30:00SPEAKERAnyway, anyway.

30:01SPEAKER- What's the happy bow for this story, Brad?

30:03SPEAKER- I don't know that there is one.

30:04SPEAKERA lot of multifaceted ways that you are getting

30:07SPEAKERa million spam text messages or scam, I should say,

30:10SPEAKERtext messages either through coordinated,

30:13SPEAKERlike very organized cyber crime

30:15SPEAKERor people just hijacking the thing hooked up

30:18SPEAKERto the traffic light down the street.

30:20SPEAKER- There was an interesting post at the end

30:22SPEAKERof the ARS article about the phishing,

30:25SPEAKERthe attacked devices from one ARS user, Eustaceus Zephyr,

30:33SPEAKERwhich says, "Why does cellular service provided

30:35SPEAKER"need to have SMS enabled?"

30:37SPEAKERAnd they said, "I created an ARS account solely

30:39SPEAKER"to comment on this.

30:40SPEAKER"I've been along, SMS is an integral part

30:42SPEAKER"of mobile connectivity.

30:43SPEAKER"It's effectively embedded everywhere

30:45SPEAKER"and cannot be removed because it would require

30:47SPEAKER"a network overhaul of revolutionary proportions

30:49SPEAKER"for that to happen.

30:51SPEAKER"It was born in 1G and 2G, persevered in 3 and 4G

30:53SPEAKER"all the way up to 5.

30:55SPEAKER"And that's why it's hard to let go of,

30:57SPEAKER"even if people don't use it for P2P communication anymore.

31:00SPEAKER"It's insecure because it's old

31:02SPEAKER"and essentially un-upgradable,

31:03SPEAKER"and that makes it an easy choice for a lot of services

31:06SPEAKER"from mass alerts and things like hurricane season

31:08SPEAKER"or delivery information, or the best-known case,

31:12SPEAKER"which is one-time password delivery.

31:14SPEAKER"One of these cases is also M2M,

31:16SPEAKER"a precursor of sorts for IoT.

31:19SPEAKER"Whether these devices in their original uses use SMS

31:21SPEAKER"for A2P alerts, manifesting as texts, or A2A

31:25SPEAKER"for automatic management, et cetera."

31:28SPEAKERLike, your phone is using SMS messages that you don't see

31:32SPEAKERto do things like deliver microcode updates to the radio

31:35SPEAKERand stuff like that.

31:35SPEAKER- Oh, interesting.

31:36SPEAKER- Yeah, so they can't just turn off SMS is the answer.

31:40SPEAKER- That makes sense.

31:40SPEAKER- Yeah.

31:41SPEAKER- You would wonder with some of this unsecured public

31:44SPEAKERor random infrastructure out there,

31:46SPEAKERif they would not be able to notice a massive uptick

31:48SPEAKERin SMS activity in and out of their stuff,

31:52SPEAKERbut if they're not paying enough attention to update

31:53SPEAKERto keep the firmware up to date,

31:56SPEAKERthey're probably also not paying enough attention

31:57SPEAKERto monitor the cellular traffic, I assume.

32:01SPEAKER- I'm honestly surprised that the carriers

32:02SPEAKERaren't paying more attention,

32:03SPEAKERbut I wonder if the networks are such a hodgepodge

32:06SPEAKERof core Verizon traffic and government-contracted traffic

32:11SPEAKERand virtual provider traffic and all that,

32:15SPEAKERthat they don't actually have a good way to track

32:17SPEAKERwhat's going on on the nodes at any given time.

32:19SPEAKER- Is this a budding example of humanity losing control

32:23SPEAKERof one of these systems it has built

32:25SPEAKERto underpin modern society?

32:27SPEAKER- Look, I watched a "Mission Impossible" the other day

32:30SPEAKERabout a rogue AI, and this is how it starts, man.

32:33SPEAKER- Yeah, I'm thinking more like the time machine maybe,

32:35SPEAKERbut we'll see.

32:36SPEAKER- Yeah, yeah.

32:39SPEAKERLet's talk about why the iRobot founder

32:43SPEAKERis wary of humanoid robots.

32:46SPEAKER- Yeah, I feel like robotics is something

32:47SPEAKERthat we are going to have to talk about a lot more

32:51SPEAKERin the relatively near, by near,

32:52SPEAKERI mean like five to 10 years, probably.

32:54SPEAKERBut I think I felt like the NVIDIA keynote at CES this year

32:58SPEAKERwas kind of the bellwether for me,

33:00SPEAKERthe amount of kind of bipedal or humanoid robots

33:03SPEAKERthat they talked about their machine learning, driving,

33:05SPEAKERor being built to drive in the future.

33:08SPEAKER- It's been a common thread in Jensen's keynotes

33:12SPEAKERfor the last couple of years now with the Omniverse stuff.

33:15SPEAKERBut yeah, it's interesting 'cause that actually ties

33:18SPEAKERinto kind of where they are.

33:20SPEAKERSo this is based on a blog post by Rodney Brooks,

33:23SPEAKERwho was one of the co-founders or one of the founders

33:25SPEAKERof iRobot, the Roomba people is what you know them as.

33:30SPEAKER- I was not familiar with this guy,

33:32SPEAKERbut he's like this classic academic turned industrialist,

33:35SPEAKERwhich I feel like you don't see a lot of these days.

33:37SPEAKER- Yeah.

33:38SPEAKER- I think that's how a lot of the like big semiconductor

33:41SPEAKERcompanies of the mid 20th century came about, right?

33:44SPEAKEROr I'm trying to figure out more recent examples.

33:46SPEAKER- And those guys were nerds who made a bunch of money, yeah.

33:50SPEAKER- Or like Sun Microsystems is a good example

33:53SPEAKERof like a bunch of those guys were writing Unix stuff

33:55SPEAKERin academic university settings or at AT&T.

33:59SPEAKERAnd then they went off and founded a company

34:01SPEAKERthat was worth many billions, you know what I mean?

34:02SPEAKERIt's just like this classic like academic to a guy

34:07SPEAKERwho founds a company and changes the world

34:08SPEAKERand makes a bajillion dollars type of pipeline.

34:11SPEAKER- As opposed to the other traditional path,

34:13SPEAKERwhich is that you were the roommate of a smart nerd

34:16SPEAKERand then you sweated your way into the CEO position

34:18SPEAKEReventually when he wanted to retire.

34:20SPEAKERNow you own a basketball team.

34:21SPEAKER- Yes, and you happen to be very loud

34:23SPEAKERand a pretty good salesman, let's say.

34:25SPEAKER- I've reached a point now with Microsoft

34:27SPEAKERthat I look back on the developers, developers,

34:28SPEAKERdevelopers days with fondness, so.

34:30SPEAKER- Yeah, kind of, sure.

34:31SPEAKER- Good job, bomber.

34:32SPEAKERLike when you're pining for the bomber days,

34:35SPEAKERyou know things are great.

34:35SPEAKERBut anyway, the blog post is really, really long.

34:38SPEAKERIt's super fascinating.

34:39SPEAKERIt's definitely worth a read if you're into humanoid robots.

34:42SPEAKERAnd he kind of, I kind of read it as the antithesis

34:45SPEAKERof Musk's, hey, make my stock price go higher,

34:49SPEAKERfueled optimism, 'cause you know, Tesla's been pushing,

34:53SPEAKERhe's been pushing humanoid robots at Tesla for a minute

34:56SPEAKERas a, I think he said a $30 trillion potential industry,

35:00SPEAKERwhich is.

35:00SPEAKER- Yeah, I bet he did.

35:01SPEAKER- Yeah, of course.

35:03SPEAKER- The site's another company called Figure

35:04SPEAKERthat I'm not familiar with.

35:06SPEAKER- Oh yeah, they make humanoid robots.

35:08SPEAKERThey make the little guys, they make the small scale ones.

35:10SPEAKER- Okay, their figure.ai is their, wow boy,

35:13SPEAKERtheir webpage sure is just a giant splash video

35:16SPEAKERof humanoid robots doing dishes and folding laundry

35:20SPEAKERand watering plants and building cars.

35:23SPEAKER- Yeah, their point, the point of this is

35:27SPEAKERthere are two potential robot assisting humans outcomes.

35:31SPEAKEROne where we have a bunch of specialized robots

35:33SPEAKERthat do one thing.

35:34SPEAKERLike you have a machine in your closet in your bedroom

35:37SPEAKERthat you put your laundry hamper of clean clothes on

35:40SPEAKERand then it spends the next six hours folding those

35:43SPEAKERand putting them in a nice stack

35:44SPEAKERso you can put them in the drawer, right?

35:46SPEAKERAnd you don't have to fold your laundry anymore.

35:48SPEAKERThink of it as like a dishwasher, but for every task,

35:51SPEAKERincluding also loading the dishwasher.

35:54SPEAKERThe other is that we just have one human-sized robot

35:58SPEAKERthat can move and walk like humans do.

36:01SPEAKERAnd this, so even like 10 years ago,

36:05SPEAKERNorman, Joey and I went with Tested

36:06SPEAKERto the DARPA Robotics Challenge in LA

36:09SPEAKERwhere they did the, like a humanoid robot,

36:12SPEAKERor actually it was a disaster recovery

36:15SPEAKERand response robot challenge,

36:17SPEAKERwhere basically the robots had to mimic the conditions

36:22SPEAKERthat the human, the people that went into Fukushima

36:25SPEAKERto like turn off the reactor and make it safe

36:28SPEAKERhad to go through and all the steps that they had to do

36:32SPEAKERin order to turn off the reactor when the tsunami happened.

36:36SPEAKERAnd because those people were exposed

36:38SPEAKERto a fair amount of radiation, it was unsafe.

36:40SPEAKERAnd they wanted to see if the challenge was,

36:43SPEAKERcan we build robots that can perform these same steps,

36:45SPEAKERget in and out, do the thing that needs to be done,

36:48SPEAKERand then egress safely so that we don't have to put humans

36:53SPEAKERin these dangerous situations.

36:54SPEAKERSo a lot of them were like university teams

36:56SPEAKERthat were bringing in Boston Dynamics Atlas robots,

36:59SPEAKERwhich are the big like eight foot tall bipedal robots.

37:04SPEAKERThey used to have a cage that-

37:05SPEAKER- I'm sorry, did you say eight foot tall?

37:07SPEAKER- They were very large.

37:08SPEAKER- Why would you build a robot that big?

37:09SPEAKERThat's terrifying.

37:10SPEAKER- Well, it was-

37:11SPEAKER- Or did it have to be at that point?

37:12SPEAKERWere they not able to miniaturize the internals enough

37:15SPEAKERto make it smaller?

37:16SPEAKER- They were electro hydraulic was how the movement all worked

37:22SPEAKERso they had like hydraulics for the legs

37:24SPEAKERand I think the arms and then servos and motors

37:28SPEAKERfor the pinchers and hands and stuff like that.

37:31SPEAKERThey were very, very large and they were,

37:35SPEAKERthere were all sorts of rules that we had to follow

37:37SPEAKERwhen we were there about being around them

37:38SPEAKERwhen they were active.

37:39SPEAKERWhen they weren't active, they had to be hung

37:40SPEAKERfrom a gantry so that if it tipped over

37:43SPEAKERor somebody bumped into it and it fell

37:44SPEAKERwhen it wasn't turned on,

37:45SPEAKERyou wouldn't get crushed to death under it.

37:48SPEAKERLike they were essentially industrial robots

37:49SPEAKERthat were shaped kind of like people and could,

37:53SPEAKERthe idea was that they were shorter than a doorway opening.

37:57SPEAKERSo maybe it was seven and a half feet tall,

37:58SPEAKERbut they were tall, they were taller enough than me

38:00SPEAKERthat I was like, oh, this is what, you know,

38:04SPEAKERit was like looking up at Ben Pack basically.

38:06SPEAKER- Sure.

38:08SPEAKER- So his whole thing about this is

38:11SPEAKERI don't wanna be anywhere near a bipedal human-sized robot

38:15SPEAKER'cause they're way more mass than a human has

38:18SPEAKERand they have a ton of stored up energy

38:20SPEAKERfrom balancing this big, giant, heavy robot

38:22SPEAKERso that they're fine when you're just walking around

38:25SPEAKERand they're doing fine, but as soon as something goes wrong

38:26SPEAKERand starts to tip over and fall and it doesn't recover,

38:29SPEAKERit's going to really, really hurt people that are around it

38:32SPEAKERor destroy whatever furniture it lands on or whatever.

38:35SPEAKER- Yeah, it was one of the big takeaways for me.

38:37SPEAKERI mean, this is a physics-y way to look at this,

38:39SPEAKERtalking about stored energy in the limbs or whatever.

38:42SPEAKERI mean, I think really what they're just saying is

38:44SPEAKERit's gonna apply a lot of force when it,

38:46SPEAKERbut my understanding of this is

38:47SPEAKERhe's not talking about it falling and landing on somebody,

38:49SPEAKERhe's talking about the robot's adaptation to falling

38:53SPEAKERto restore its balance, right?

38:55SPEAKERIt's talking about, you can imagine how rapidly the limbs

38:58SPEAKERwould move to prevent the robot from falling in mid fall

39:02SPEAKERand if somebody is caught in that,

39:04SPEAKERif somebody is struck by a leg or arm that is shooting out

39:07SPEAKERto stop the robot from falling,

39:09SPEAKERthat seems like very, very bad injury territory.

39:13SPEAKER- And in the paper, in the blog post,

39:14SPEAKERhe talks about the mass differential

39:16SPEAKERbetween a human meat and bone and a robot arm,

39:20SPEAKERand it's basically 4X.

39:21SPEAKERSo if your arm weighs 10 pounds, you fling your arm out

39:24SPEAKERand the small light end of your arm in your hand,

39:26SPEAKERwhich weighs a pound, hits somebody,

39:29SPEAKERthe arm that weighs on the robot will weigh 40 or 50 pounds

39:32SPEAKERand the hand will weigh five or 10,

39:33SPEAKERand it's like getting hit by a brick in your face, right?

39:36SPEAKERSo, you know, and like I've been in car factories

39:43SPEAKERwhere they have industrial robots

39:44SPEAKERand one of the defining characteristics of that experience

39:49SPEAKERis that there are human safe areas of the factory

39:51SPEAKERand then there's areas of the factory

39:53SPEAKERwhere the industrial robot is,

39:55SPEAKERand you don't go into the industrial robot,

39:58SPEAKERlike there's typically multiple,

40:00SPEAKERlike it takes one person watching on the controls,

40:03SPEAKERpressing a button to open the doors to go into the areas,

40:06SPEAKERthere's like motion sensors that stop the robot

40:08SPEAKERdoing whatever it's supposed to be doing

40:10SPEAKERas soon as somebody's in the space

40:11SPEAKERthat's not supposed to be there,

40:13SPEAKERto the point that like if something rolls into an area,

40:17SPEAKERone of the no-go zones, it'll stop the whole line

40:19SPEAKERbecause all the robots stop working, right?

40:23SPEAKERAnd like I hadn't really thought about it in this context,

40:26SPEAKERbut I do remember distinctly walking around

40:28SPEAKERthat DARPA challenge and it being kind of unsettling.

40:31SPEAKERNow, there were a bunch of other kinds of robots.

40:34SPEAKERThere was JPL had an octopad or a hexapod, I guess,

40:39SPEAKERa six or eight legged robot that kind of scuttered in.

40:42SPEAKERThere were some wheeled robots,

40:44SPEAKERthere were some tracked stuff,

40:46SPEAKERthere was all sorts of different stuff there.

40:47SPEAKERWe did a couple of videos there,

40:48SPEAKERthey're still on the tested channel

40:49SPEAKERif you wanna go find them.

40:51SPEAKERThe other thing he talks about a lot

40:53SPEAKERis that we're kind of not training today's robots

40:58SPEAKERto be dexterous.

40:59SPEAKERThis is the thing that was the actual,

41:01SPEAKERI mean, the thing we just described is brightening,

41:03SPEAKERbut something that can be accounted for

41:05SPEAKERwith enough safety precautions and proper training

41:07SPEAKERfor the robots in place.

41:09SPEAKERIt was the very slapdash way

41:11SPEAKERthese things are being trained.

41:12SPEAKERThat was the like, oh, we are fucking this up bad.

41:14SPEAKERThis is like Tesla's laissez-faire attitude

41:17SPEAKERtowards self-driving times 10.

41:19SPEAKER- Well, so, okay, so there's two things.

41:22SPEAKEROne is that the dexterity,

41:24SPEAKERwe don't have robot equivalents of fingers.

41:27SPEAKERWe have grippers that are just like

41:29SPEAKERslide-in, slide-out kind of pliers

41:32SPEAKERthat are on a linear rail, basically,

41:33SPEAKERso they just open and close,

41:35SPEAKERbut those aren't able to do something

41:37SPEAKERlike pick up a match or a pencil

41:39SPEAKERor something like that, necessarily.

41:41SPEAKERThen we have suction,

41:42SPEAKERwhich is used a lot in industrial situations

41:43SPEAKERwhere you're like, pick-and-place machines

41:45SPEAKERoften use suction in an assembly context for electronics.

41:53SPEAKERAs an example of that DARPA challenge,

41:55SPEAKERthe most common way that the robots opened a door

42:00SPEAKERwas just by punching,

42:01SPEAKERa couple of them just literally

42:03SPEAKERpunched through the part of the door

42:05SPEAKERthat held the door closed

42:07SPEAKERand then yanked it back and opened the door

42:09SPEAKERwithout twisting the knob or anything

42:11SPEAKER'cause they couldn't work,

42:12SPEAKERthey couldn't identify and work the knobs autonomously.

42:15SPEAKERHe specifically cites 15 tasks

42:18SPEAKERthat they've come up with

42:18SPEAKERthat an eight-year-old human can do

42:20SPEAKERfor like a how are we doing Olympiad for robot dexterity,

42:24SPEAKERand it's stuff like folding laundry

42:26SPEAKERwhere one sleeve is inside out

42:28SPEAKERand you need a button button

42:29SPEAKERto have the shirt folded properly,

42:32SPEAKERcleaning peanut butter off its own hand,

42:35SPEAKERstuff that any human can do really, really easily,

42:39SPEAKERbut that there's no robot equivalent

42:41SPEAKERand a robot to clean its own hand

42:43SPEAKERis an unthinkable task right now.

42:46SPEAKER- Yeah, the biggest thing in this

42:47SPEAKERthat seems incredibly obvious once you read it

42:50SPEAKERbut you don't really think about upfront

42:53SPEAKERis how incredibly dependent on the sense of touch humans are

42:56SPEAKERfor guiding their motion and dexterity

42:59SPEAKERand that we're not building robots

43:01SPEAKERwith any kind of sense of touch

43:03SPEAKERor like touch feedback mechanism at all.

43:06SPEAKER- Well, it's not just on the,

43:07SPEAKERso two things about that.

43:08SPEAKEROne is that even the robots that we do have

43:11SPEAKERthat have finger equivalents

43:13SPEAKERand do have touch on the fingertips

43:15SPEAKERaren't getting feedback from things like your wrist

43:18SPEAKERor the other parts of the arm, wrist and elbow as much.

43:22SPEAKERThe other thing I didn't realize

43:24SPEAKERis that those hand analog controls are not durable at all.

43:28SPEAKERI didn't realize the kind of number of cycles

43:31SPEAKERthat those things can go through.

43:32SPEAKERHe talks about it a little bit,

43:33SPEAKERbut he's like, there's a reason everything does grippers

43:36SPEAKERinstead of fingers

43:37SPEAKERand it's 'cause the fingers wear out incredibly fast

43:39SPEAKERin the robot form.

43:40SPEAKER- So on the one hand, like that's a huge amount of data

43:42SPEAKERthat humans depend on to guide the use of their hands

43:46SPEAKERthat robots are barely, if at all, getting.

43:48SPEAKERThe other thing is that Tesla and Figure

43:50SPEAKERare the two companies being cited here.

43:52SPEAKERI don't know if this could vary for other startups

43:54SPEAKERthat might be working on this.

43:55SPEAKERIt sounds like they're just using pure video

43:57SPEAKERto train robots' behavior at this point

44:00SPEAKERand that they have actually, in fact,

44:02SPEAKEReven pulled back from more robust sources of data

44:05SPEAKERto feed into them pretty recently.

44:08SPEAKER- This is happening in more than just the humanoid robotics.

44:11SPEAKERThere's a lot of car self-driving startups.

44:14SPEAKER- That was another big thing at CES

44:15SPEAKERin the video keynote was the,

44:17SPEAKER'cause that was the big venue for synthetic test data,

44:21SPEAKERremember, that they were simulating a bunch of driving data

44:24SPEAKERto feed into the model.

44:25SPEAKER- Yeah, and there's, I've talked to some people

44:29SPEAKERwho work in those spaces 'cause it's one of the places

44:31SPEAKERthat game developers who don't wanna make games

44:33SPEAKERend up going anymore is they build fake worlds

44:35SPEAKERthat the self-driving cars can learn to drive in

44:38SPEAKERand it's easy to replicate.

44:40SPEAKERIt makes it easy to create millions of miles

44:44SPEAKERor tens of millions of miles, the kind of training

44:47SPEAKERthat Waymo has been doing for the last 10 years

44:49SPEAKERon the streets of San Francisco and the kind of data set

44:52SPEAKERthat Tesla has been generating for the last 15 years

44:56SPEAKERby essentially charging people

44:58SPEAKERto be a giant data collection machine,

45:00SPEAKERto drive a giant data collection machine.

45:05SPEAKERMy understanding is that for cars,

45:08SPEAKERwhich is a system where there's a fair number,

45:11SPEAKERthe attractable number of rules

45:13SPEAKERthat are fairly well established

45:15SPEAKERand it's a fairly standard problem

45:19SPEAKERin that we know the edges and we know where the edge cases

45:21SPEAKERare and all that stuff, it works reasonably well

45:25SPEAKERup to a point.

45:26SPEAKERIt'll be the first 20 million miles of train data

45:29SPEAKERthat you need and then you need another 10 million

45:31SPEAKERbeyond that on real roads.

45:33SPEAKERNow, houses, as he said in this blog post,

45:38SPEAKERare wildly different.

45:40SPEAKEREven a really tidy house is gonna be a nightmare

45:44SPEAKERand your house and my house and everybody listening

45:46SPEAKERto this houses are all different.

45:48SPEAKERWe all have different expectations for things.

45:50SPEAKERWe all have different levels of complexity in our homes

45:53SPEAKERand it's a much more thorny problem.

45:57SPEAKER- Yeah, and even if the home as a physical space

46:00SPEAKERis knowable at some point, but there's an infinite

46:03SPEAKERcombination of configurations of human beings

46:05SPEAKERand scenarios that can exist within the home

46:07SPEAKERat any given time.

46:08SPEAKER- Well, and just think about the product diversity

46:10SPEAKERof things you find in your home in terms of furniture

46:13SPEAKERand you don't see very many hundred year old cars

46:18SPEAKERon the road, so the machines can recognize the most,

46:23SPEAKERif they recognize the thousand most common cars,

46:25SPEAKERthat'll be 90% of all the cars that they ever see

46:27SPEAKERin the life of the driving.

46:29SPEAKERBut in your house, you can have furniture

46:31SPEAKERthat's a hundred years old.

46:32SPEAKERYou can have cribs and chairs and divans

46:36SPEAKERand all these millions of different kinds of furniture

46:39SPEAKERand items on shelves and all this stuff

46:42SPEAKERand it's not, it's a noisy environment.

46:46SPEAKER- The other thing he talks about is the speed

46:50SPEAKERat which it takes the robots to do something.

46:52SPEAKERSo the example there, he talks about a robot

46:55SPEAKERtaking 30 seconds to light a match

46:56SPEAKERwhere a human takes seven seconds.

46:59SPEAKERI have some friends who are working

47:00SPEAKERin the laundry folding robots and their goal

47:03SPEAKERis to have a machine that will fold a basket of laundry

47:08SPEAKERover the course of a day, not in 10 minutes or 20 minutes,

47:13SPEAKERhowever long it would take a person to fold

47:14SPEAKERthe equivalent amount of laundry.

47:16SPEAKERThey kind of don't care if it takes a long time

47:17SPEAKER'cause it's a passive experience for you, the user.

47:20SPEAKERIt was, it's a, like I said, fascinating blog post.

47:25SPEAKERHe talks about building robots that work in warehouses

47:28SPEAKERwhich are fully controlled environments in those cases.

47:32SPEAKERThere's a big section about hands

47:34SPEAKERthat we just barely touched on and the machine,

47:37SPEAKERlike this is one of the uses of machine learning

47:39SPEAKERthat I think is actually really interesting

47:41SPEAKERand valid and valuable but figuring out how to train it

47:44SPEAKERseems really challenging.

47:46SPEAKER- Yeah, so back, I mean, back to the training thing,

47:48SPEAKERthere's just a couple of gaps to fill in there.

47:50SPEAKERWe're talking about this in the context

47:51SPEAKERof these companies training robots to do human jobs

47:54SPEAKERand of course they need real people doing those jobs

47:56SPEAKERto provide data on how to do the jobs

47:58SPEAKERto provide to the robots.

47:59SPEAKERPreviously, it sounds like Tesla

48:01SPEAKERwas using motion capture suits.

48:03SPEAKERSo you're at least getting motion data

48:04SPEAKERof a real person doing a job.

48:06SPEAKERAnd teleoperation is the other term they use here.

48:08SPEAKERI assume that means a remote operator like driving the robot.

48:12SPEAKER- Yeah, so that's how like Waymo cars, for example,

48:14SPEAKERwhen they get stuck and the robot doesn't know what to do,

48:16SPEAKERthen somebody in a call center someplace logs in,

48:21SPEAKERsees what's going on and gives it instructions.

48:23SPEAKER- Right, so they were using both people

48:25SPEAKERin motion capture suits doing the work

48:26SPEAKERand also teleoperators, I guess,

48:28SPEAKERdriving the robots remotely doing the work

48:30SPEAKERas additional sources of data

48:31SPEAKERfor training the robots on doing the work.

48:34SPEAKERBut now they have shifted to purely video,

48:36SPEAKERpresumably internal and external.

48:39SPEAKERFrom the worker's perspective or maybe both,

48:41SPEAKERmaybe third person as well.

48:43SPEAKERUsually it's backpack plus a third person,

48:45SPEAKERthird person external cameras.

48:47SPEAKER- Yeah, that sounds questionable to me.

48:49SPEAKERI mean, and this is just to be clear,

48:51SPEAKERin case anybody's like yelling at the radio,

48:53SPEAKERthis is completely separate from the issue of training robots

48:56SPEAKERto do people's jobs at a time when people

48:58SPEAKERare already getting automated out of work rapidly.

49:02SPEAKERThis is like leading to a level of dystopia

49:04SPEAKERthat's kind of hard to wrap your head around

49:06SPEAKERin socioeconomic terms, but that's a whole other subject.

49:09SPEAKERLike really, I'm just talking about the safety aspect here

49:12SPEAKERof teaching the robots who behave safely

49:15SPEAKERand responsibly around people with a very limited data set.

49:18SPEAKER- Well, so there's a couple of things.

49:20SPEAKEROne is that the fidelity and resolution of the data

49:24SPEAKERthat comes from the motion capture suits isn't great.

49:27SPEAKERWe're still not good at tracking fingers, frankly.

49:30SPEAKERSo for a dexterous task like taking, say,

49:33SPEAKERsilverware out of your dishwasher

49:35SPEAKERand putting it in the appropriate slots,

49:38SPEAKERthe actual picking up and grabbing the silverware

49:41SPEAKERis something that you can't learn

49:44SPEAKERin the way that you would expect.

49:46SPEAKERSame thing for folding clothes.

49:47SPEAKERIt's a similar problem 'cause you have to hold

49:49SPEAKERand manipulate the clothing in a bunch of different ways

49:52SPEAKERto figure out where the shoulders and arms and buttons

49:56SPEAKERand all that stuff on the shirt are,

49:57SPEAKERwhether it's right side out or inside out, stuff like that.

50:01SPEAKEROr balling up a pair of socks is a finger task

50:04SPEAKERthat you can get good data from.

50:05SPEAKERSo that's one of the reasons they switched to cameras.

50:08SPEAKERI know, I don't know what the right solution is for this.

50:15SPEAKERI think you have to do a combination of everything.

50:17SPEAKERAnd it seems like dogmatic approaches

50:20SPEAKERare probably not the right ones

50:22SPEAKERbecause you need to do the right training

50:25SPEAKERfor the right solution, for the right situation.

50:27SPEAKERAnd sometimes it's gonna be cameras.

50:28SPEAKERSometimes it's gonna be motion capture.

50:29SPEAKERSometimes it's gonna be depth cameras.

50:31SPEAKERSometimes it's gonna be all of the above.

50:34SPEAKERIt's wild.

50:35SPEAKERI would not have expected this to come from the Roomba guy.

50:38SPEAKERAlthough I think the Roomba is a good example

50:41SPEAKERof we built a thing to work inside the limitations

50:45SPEAKERof a specific kind of device that's pretty good.

50:48SPEAKERAnd yeah.

50:52SPEAKER- Is it this R story that says this guy

50:55SPEAKERhas kind of shied away from humanoid robot development?

50:58SPEAKERIn general?

50:59SPEAKER- He specifically talks in the blog post.

51:03SPEAKERI skipped straight to the blog post.

51:04SPEAKERI didn't read the R story.

51:06SPEAKERBut he skipped straight to the blog post

51:08SPEAKERabout saying that they identified the problems

51:10SPEAKERwith humanoid robots as being eternally 10 years off.

51:16SPEAKERBecause it was much easier to make single purpose robots

51:20SPEAKERor robots for controlled areas like factories and warehouses

51:24SPEAKERthan it is for the chaotic environment that humans live in.

51:27SPEAKER- That's exactly the point that the R story ends on.

51:31SPEAKERBrooks predicts that within 15 years

51:32SPEAKERthere will need many robots called humanoids,

51:35SPEAKERbut ironically they will look nothing

51:36SPEAKERlike today's bipedal machines.

51:38SPEAKERThey'll have wheels instead of feet,

51:39SPEAKERvarying numbers of arms and specialized sensors

51:42SPEAKERthat bear no resemblance to human eyes.

51:44SPEAKER- That's it.

51:45SPEAKERThat makes sense to me.

51:46SPEAKER- Yeah, some will have cameras in their hands

51:48SPEAKERor looking down from their midsections, et cetera.

51:51SPEAKER- Rosie the robot will be a team of helper robots

51:55SPEAKERand probably will take,

51:57SPEAKERthen instead of washing our dishes and folding our laundry,

52:00SPEAKERwe'll spend three hours a week in robot maintenance,

52:03SPEAKERI'm sure.

52:04SPEAKER- I don't know also who is the us in this scenario

52:06SPEAKERbecause this--

52:07SPEAKER- You're not buying a robot?

52:08SPEAKER- That sounds like the kind of thing

52:09SPEAKERthat is going to be geared toward the very ultra wealthy,

52:12SPEAKERbut I don't know.

52:13SPEAKER- Well, you know, buy that lottery ticket, Brad.

52:17SPEAKERSpeaking of things that the very ultra wealthy

52:20SPEAKERaren't gonna care about,

52:21SPEAKERWindows 10 end of life a couple of weeks from now.

52:23SPEAKER- That's right, 64 bit processors.

52:25SPEAKER- Yep.

52:26SPEAKER- Not accessible to most people.

52:28SPEAKER- No, look, if you wanna run that 32 bit code

52:32SPEAKERon a native 32 bit OS, I got bad news for you

52:34SPEAKER'cause it's all going away.

52:36SPEAKERI hadn't realized, but I mean,

52:38SPEAKERI guess I had realized at some point,

52:39SPEAKERbut Windows 10 is the last version of Windows

52:42SPEAKERthat shipped with a 32 bit option.

52:44SPEAKER- Yeah.

52:46SPEAKER- I went and tried to find the last 32 bit processor

52:49SPEAKERthat Intel sold and I couldn't--

52:51SPEAKER- Oh wait, don't tell me, don't tell me.

52:54SPEAKERProbably 2007 or eight maybe?

52:58SPEAKER- Couldn't figure out which one it was.

53:00SPEAKERI had to guess, I'm pretty sure the last 32 bit CPU I bought

53:04SPEAKERwas around 2006, which was an Athlon XP.

53:07SPEAKERIt was like that was the last,

53:08SPEAKERit was also the last single core CPU I bought.

53:11SPEAKERI don't know, I don't know if they went to 64 bit

53:13SPEAKERright around the same time

53:14SPEAKERthat multiple cores started becoming a thing, but.

53:17SPEAKER- So I thought it was gonna be,

53:19SPEAKERI thought the last thing that would have been mainstream

53:21SPEAKERwould have been an Atom processor.

53:24SPEAKER- Oh yeah, if you're talking like kind of low end

53:26SPEAKERand embedded, like I'm sure they were shipping 32 bit only

53:29SPEAKERfor a long time past desktop.

53:31SPEAKER- I think that there were a lot of e-machines

53:33SPEAKERthat used Atom, or sorry, e-books.

53:36SPEAKERWhat are they called, netbooks that used Atom,

53:39SPEAKER32 bit Atom processors is the last thing that I could find

53:41SPEAKERthat was like anything mainstream.

53:43SPEAKERBut I don't know what the last 32 bit only,

53:45SPEAKERI couldn't figure out what the last 32 bit only processor

53:47SPEAKERwas 'cause it was kind of ambiguous.

53:50SPEAKERLike I wanted to, it's pretty easy to find the launch dates,

53:53SPEAKERbut it's hard to find like end of sale dates.

53:56SPEAKERSo anyway, if you know, please write in,

53:58SPEAKERI would love to know techbot@content.town.

54:02SPEAKERBut, so 32 bit support with Windows 10 goes away

54:06SPEAKERon October 24th when they end of life Windows 10.

54:10SPEAKERWindows 11 didn't ship with 32 bit version at all.

54:13SPEAKERI think people have hacked them together or something.

54:16SPEAKER- Not really.

54:17SPEAKER- I'm sure, like people do all sorts of crazy stuff

54:19SPEAKERthat they shouldn't do with Windows these days.

54:21SPEAKER- Wait, how would you do that?

54:22SPEAKERI mean, like kernel would have to be recompiled

54:24SPEAKERfor that, right?

54:25SPEAKER- I think that they're probably pulling old versions

54:27SPEAKERof kernels and porting them forward, I don't know.

54:30SPEAKERWhen I Googled it, I found stuff that I looked at

54:32SPEAKERand I was like, oh, this is a bad idea,

54:33SPEAKERI'm not gonna click into this link.

54:36SPEAKERI wanted to talk a little bit

54:37SPEAKERabout the 64 bit timeline though, 'cause--

54:40SPEAKER- Okay.

54:40SPEAKER- Microsoft shipped XP AMD 64 edition officially in 2005,

54:45SPEAKERbut there was a beta that was available publicly

54:48SPEAKERas early as 2003 is my understanding.

54:52SPEAKERVista shipped with 64 bit versions on disks

54:55SPEAKERfor most of the retail versions, I think Ultimate and Pro,

54:59SPEAKERI think maybe on Home you didn't get one or something,

55:01SPEAKERthere was some division based on the SKU that you purchased.

55:04SPEAKERAnd PCs that had 64 bit processors shipped with support

55:08SPEAKERturned on out of the box,

55:09SPEAKERwhich was like a big deal at the time.

55:13SPEAKERFor folks who don't remember, I think if 64 bit

55:15SPEAKERlet us go beyond four gigs of memory,

55:17SPEAKERI think was the big jump.

55:19SPEAKER- Was it four gig?

55:21SPEAKER- I think it was four gigs.

55:23SPEAKERAnd seven defaulted to 64 bit unless you specified 32 bits.

55:28SPEAKERNow, the important thing to note for all these

55:33SPEAKERback in the day, it doesn't matter anymore,

55:35SPEAKERis that for most of these to make the 32 to 64 bit jump,

55:38SPEAKERyou had to go, you had to do a clean install,

55:40SPEAKERexcept for the XP to Vista install,

55:45SPEAKERwhich would do a 64 bit upgrade I think for you.

55:48SPEAKER- Sure, it was four gigabytes, I just went and looked,

55:50SPEAKERI used to know things like that.

55:52SPEAKER- This hasn't been important for 20 years.

55:53SPEAKERIt's fine to forget that.

55:54SPEAKER- Fair, that's fair.

55:57SPEAKER- So anyway, it seems like this means

56:00SPEAKERthat 32 bit support stuff is fading in general,

56:02SPEAKER'cause Valve announced that Steam's gonna drop support

56:05SPEAKERfor 32 bit Windows systems in January of 2026.

56:09SPEAKER- Yeah, 32 bit client of Steam to be clear.

56:11SPEAKERLike that's just the actual Steam client

56:13SPEAKERyou use to download and launch games.

56:15SPEAKERIt sounds like any 32 bit games that you have

56:18SPEAKERon a 64 bit version of Steam are still fine to run

56:21SPEAKERand presumably are supported.

56:23SPEAKER- Well, I mean, support for those games

56:24SPEAKERis down to their developers, right?

56:26SPEAKER- Yeah, well, and Firefox is gonna stop shipping

56:29SPEAKER32 bit versions of Firefox for Linux,

56:32SPEAKERI think is what they said.

56:34SPEAKERI couldn't find out whether they're still shipping

56:36SPEAKER32 bit Windows versions of Firefox.

56:40SPEAKERThat didn't seem to be readily available.

56:42SPEAKERLinux kernels are starting to drop support

56:43SPEAKERfor 32 bit processors now, which-

56:45SPEAKER- Yeah, I think that may have been ongoing for a while.

56:48SPEAKERActually, the thing I can say,

56:49SPEAKERbecause Debian Trixie just came out and Debian,

56:51SPEAKERespecially Debian Stable is very highly conservative

56:54SPEAKERabout this stuff.

56:55SPEAKERThey have not like 1,000% deprecated 32 bit support

56:59SPEAKERwith Trixie, which came out about two months ago,

57:02SPEAKERnot quite two months ago.

57:03SPEAKER- Okay.

57:04SPEAKER- But they almost have, like they have all,

57:06SPEAKERbut I think the only thing you can do now

57:09SPEAKERis install 32 bit packages on a 64 bit install.

57:12SPEAKERBut if you cannot, if I'm not mistaken,

57:14SPEAKERyou cannot install a fully 32 bit from the ground up

57:19SPEAKERDebian install, and again, Debian, I think is usually

57:21SPEAKERone of the last to do things like that.

57:23SPEAKER- Yeah, it seems like it.

57:24SPEAKER- So there's like a big C change.

57:26SPEAKERLike it just kind of, it's happening industry-wide.

57:29SPEAKER- Now, my first thought was does this mean

57:31SPEAKERthat Windows is finally gonna get rid

57:33SPEAKERof the stupid program files/program files

57:37SPEAKERparentheses x86 folder?

57:40SPEAKER- Well, no, 'cause that's for 32 bit applications, right?

57:42SPEAKER- Yeah, I know.

57:43SPEAKERYeah, it's still gonna haunt us.

57:45SPEAKER- Yeah.

57:46SPEAKER- Really goofy.

57:47SPEAKER- But it only haunts you if you look at your C drive.

57:50SPEAKERI just, I gave up on the C drive

57:52SPEAKERas no man's land a long time ago.

57:54SPEAKER- Wow, you don't look at the C drive?

57:55SPEAKER- I've went completely the other direction

57:57SPEAKERfrom people who get really angry that you can't open,

58:00SPEAKERlike what's the, what's the Microsoft store?

58:02SPEAKER- Microsoft store apps, yeah.

58:04SPEAKER- Like Windows apps, I forget what the term is

58:06SPEAKERfor that style of application

58:07SPEAKERthat sandboxes the install directory.

58:10SPEAKER- Windows WSL or something is the API, I can't remember.

58:14SPEAKERWell, WSL is the Linux subsystem, but anyway,

58:19SPEAKERthe point being, you know, some people,

58:21SPEAKERI know some people who like jump through hoops

58:23SPEAKERto make those directories accessible to them

58:25SPEAKERbecause they're indignant about not being able

58:27SPEAKERto look at files on their own drive,

58:29SPEAKERwhich on the one hand I get,

58:32SPEAKERon the other hand, it just feels like a losing battle.

58:36SPEAKER- W, Windows app SDK and GDK is the games one now.

58:44SPEAKER- Anyway, I just, you know, it's just-

58:47SPEAKER- It's just all Microsoft territory, huh?

58:49SPEAKER- Like just messing with the install files

58:51SPEAKERfor the operating system is just pointless at the point,

58:53SPEAKERjust let it do what it's gonna do.

58:55SPEAKER- What if you need to update your host file

58:56SPEAKERand put something in there, you know?

58:57SPEAKER- I do that, I have in fact updated the host file

59:00SPEAKERon my Windows install in the last six months.

59:02SPEAKER- There you go, me too.

59:03SPEAKERI love the host file.

59:04SPEAKER- But then I moved to, I used to like,

59:08SPEAKERI used to use that to access certain local IPs

59:10SPEAKERlike on my NAS around here,

59:11SPEAKERbut then I just moved that into my blocky install.

59:14SPEAKERSo now they work network wide.

59:16SPEAKER- Oh, so you have DNS entries for them now?

59:17SPEAKER- Yes, so now they just, any client that gets DHCP on my LAN

59:21SPEAKERgets those same addresses.

59:24SPEAKER- That was the one really nice thing

59:25SPEAKERabout using DHCP on the pie holes.

59:27SPEAKERI was able to do that years ago

59:28SPEAKERand it is incredibly convenient, it turns out.

59:31SPEAKER- It's really neat.

59:31SPEAKERIf you wanna use shoe.vps to access the address

59:36SPEAKERfor your VPS, for example, within your LAN, you can do that.

59:39SPEAKER- I will tell you-

59:40SPEAKER- Every client.

59:41SPEAKER- It works great right up until the point

59:43SPEAKERthat you have a motherboard with two network cards in,

59:46SPEAKERthey both get different IP addresses

59:47SPEAKERand then everything breaks down when you do that, so.

59:52SPEAKERBut yeah, so that's it.

59:54SPEAKERLet's see, we haven't talked about

59:55SPEAKERthe Windows 10 end of life in a minute,

59:56SPEAKERbut it still seems to be on like Donkey Kong.

59:59SPEAKERI still think they're gonna blink,

1:00:01SPEAKERbut also I'm pretty confident that nobody that remains

1:00:04SPEAKERat Microsoft was there when Windows XP

1:00:06SPEAKERcollapsed 20 years ago.

1:00:08SPEAKERAnd so maybe not, I think the routers,

1:00:12SPEAKERthe fact that everybody's putting their computers

1:00:14SPEAKERbehind a router now might actually save them

1:00:16SPEAKERbecause it'll prevent the kind of rampant worm spreading

1:00:19SPEAKERthat was happening in 2002, 2001 with XP,

1:00:23SPEAKERor 2002, 2003 with XP from happening.

1:00:28SPEAKERBut at the same time, it looks like there's a better

1:00:31SPEAKERthan even chance at the end of this month,

1:00:33SPEAKERthere's going to be a billion-ish computers in the world

1:00:36SPEAKERthat don't have security updates coming anymore.

1:00:39SPEAKERAnd I guarantee you there's a bunch of zero-day exploits

1:00:41SPEAKERthat malicious users have out there ready to go.

1:00:45SPEAKERSo we'll- - Just waiting.

1:00:46SPEAKER- Yeah, just waiting for the clock to turn over.

1:00:48SPEAKER- It feels too late to blink to me.

1:00:50SPEAKERIt feels like if they were going to blink,

1:00:51SPEAKERthey would have done it by now.

1:00:52SPEAKER- Well, they've kind of done a staggered blink

1:00:54SPEAKER'cause they did the thing where if you use Windows backup,

1:00:58SPEAKERwhich is there, backup your computer

1:00:59SPEAKERto your One Cloud folder and also pay us more

1:01:02SPEAKERfor storage every month service, you can do that.

1:01:05SPEAKERAnd that'll give you another year of XP support.

1:01:08SPEAKERAnd then they also have a thing where you can trade

1:01:10SPEAKERin your Microsoft points.

1:01:13SPEAKERI can't remember.

1:01:14SPEAKERThey have a whole loyalty program thing

1:01:17SPEAKERthat if you use Bing or if you buy stuff

1:01:21SPEAKERfrom the Microsoft store, it's called Microsoft Rewards,

1:01:24SPEAKERand it's part of Bing.

1:01:26SPEAKERYou can basically, for I think 1,000 points,

1:01:30SPEAKERyou can get an additional year of service

1:01:33SPEAKERor year of updates, which at the point

1:01:36SPEAKERthat they're doing this, I'm not as worried

1:01:39SPEAKERabout the folks who are listening to this podcast

1:01:41SPEAKERand know that you can get a year of service,

1:01:46SPEAKERyear of Windows 10 updates or whatever,

1:01:48SPEAKERas I am worried about the other 999 million people

1:01:52SPEAKERwho don't know about this stuff

1:01:54SPEAKERand are gonna have computers that are just easy targets

1:01:58SPEAKERfor malware in a month.

1:02:00SPEAKER- Yeah, that's probably true.

1:02:02SPEAKERTo be clear, you can also just pay for that extended support

1:02:04SPEAKERwhich is called Windows 10 ESU extended security updates.

1:02:08SPEAKER- Is that for normal users?

1:02:10SPEAKER- Yeah, it's 60 bucks.

1:02:10SPEAKERYou can just pay for the year of,

1:02:12SPEAKERI'm looking at the Microsoft Ignite page on it right now.

1:02:17SPEAKER- For what it's worth, we did a thing.

1:02:19SPEAKERWe tried to see how fast we could get 1,000 points

1:02:23SPEAKERon the PC world on the full nerd a few months ago,

1:02:26SPEAKERand it took about 15 minutes,

1:02:29SPEAKER'cause there's usually something that's like,

1:02:30SPEAKERhey, you wanna get 1,000 points?

1:02:32SPEAKERGo click on this link,

1:02:33SPEAKERand then you get your 1,000 points real quick.

1:02:35SPEAKER- Yeah, that's probably the low-hanging fruit.

1:02:37SPEAKERI doubt that could be replicated over and over,

1:02:39SPEAKERbut that's enough to get you a year.

1:02:41SPEAKERApparently this is running for multiple years

1:02:43SPEAKERfrom the sound of things.

1:02:45SPEAKER- The ESU is, it gets more expensive each year.

1:02:48SPEAKER- Yes.

1:02:48SPEAKER- I don't know about the Bing giveaway,

1:02:52SPEAKERthe Bing reward stuff, so.

1:02:54SPEAKER- Oh, I'm sorry, it's $60 a year through volume licensing.

1:02:57SPEAKER- Yeah, I don't think they let consumers

1:02:59SPEAKERbuy into that last I saw.

1:03:01SPEAKER- Maybe not, actually.

1:03:02SPEAKERAnd that's gonna end someday anyway,

1:03:04SPEAKERlike they're not gonna do extended support forever,

1:03:06SPEAKERlike at some point you're gonna have to move on

1:03:08SPEAKERfrom Windows 10 and that's,

1:03:10SPEAKERa year will go by pretty quickly.

1:03:12SPEAKER- The last thing they said was that they were gonna do that

1:03:14SPEAKERfor two years and the second year

1:03:15SPEAKERis gonna escalate in price, so.

1:03:17SPEAKER- Yeah, it'll be really interesting to see how this goes

1:03:19SPEAKERonce we get to the date and whatever mitigations

1:03:22SPEAKERor backtracking they might do.

1:03:25SPEAKER- I can't imagine, yeah, I don't know.

1:03:29SPEAKERIt seems bad, it seems real bad.

1:03:32SPEAKERI've never been a huge antagonist on the,

1:03:36SPEAKERhey, Microsoft is wrong for forcing people

1:03:38SPEAKERto use the upgrades, to do upgrades

1:03:41SPEAKERin the post Windows 8, Windows is free

1:03:43SPEAKERfor anybody who has an old version of Windows era.

1:03:47SPEAKERBut this one, because they're requiring the newer hardware,

1:03:51SPEAKERthey're gonna leave a whole bunch of computers

1:03:53SPEAKERon Windows 10 that have no upgrade path

1:03:55SPEAKERand that feels real bad to me.

1:03:58SPEAKER- Yeah, I mean, we've talked about it before,

1:03:59SPEAKERI'm kinda mixed on it because you do have to move on eventually

1:04:02SPEAKERand security is more and more of a thing

1:04:05SPEAKERin the computing world these days.

1:04:07SPEAKERKeeping things updated and secure is more and more important

1:04:09SPEAKERthan ever as we have seen.

1:04:11SPEAKER- I think I'd feel less bad about it

1:04:13SPEAKERif Windows 10 were better.

1:04:15SPEAKER- You mean 11?

1:04:16SPEAKER- 11 were better, rather, yeah.

1:04:17SPEAKER- Yeah, well, yes.

1:04:18SPEAKERI mean, the bigger question there, though,

1:04:20SPEAKERis was Windows 11 the time to rip off this Band-Aid

1:04:22SPEAKERor should they have given it one more Windows version

1:04:24SPEAKERbefore they forced the TPM?

1:04:26SPEAKER- Well, yeah, and then the other thing is,

1:04:29SPEAKERat this point, I remember Windows 10

1:04:30SPEAKERas it was two or three years ago, not as it is today,

1:04:33SPEAKERand I'm sure they've been shitified Windows 10

1:04:35SPEAKERat this point, too, so.

1:04:37SPEAKER- I've heard it's the year of the Linux desktop,

1:04:38SPEAKERmight be worth looking into.

1:04:40SPEAKER- Yeah, you know, wherever fine podcasts are found.

1:04:42SPEAKER- Or the FreeBSD desktop as the case may be.

1:04:45SPEAKER- Well, so that's "Double Diaries" season two?

1:04:48SPEAKER- Yes, I heard that at the end of last week's episode.

1:04:51SPEAKER- Yeah, and then "Haiku" is season three, I think.

1:04:53SPEAKER- Even I would not run a FreeBSD desktop,

1:04:55SPEAKERwhich is not to say you can't.

1:04:57SPEAKERLike, it's certainly possible, but--

1:04:59SPEAKER- My reaction to that was pretty strongly negative,

1:05:01SPEAKERif I recall.

1:05:02SPEAKER- Yeah.

1:05:03SPEAKER- Anyway, I think that's as good a place as any

1:05:05SPEAKERto wrap it up for this week.

1:05:07SPEAKERThanks, everybody, for listening.

1:05:08SPEAKERThanks for supporting the show.

1:05:10SPEAKERAs always, "Brad and Will Made a TechPod"

1:05:12SPEAKERis a listener-supported show,

1:05:14SPEAKERwhich means we wouldn't be here without you, the listeners.

1:05:17SPEAKERIf you would like to support the show,

1:05:18SPEAKERyou can go to patreon.com/techpod,

1:05:21SPEAKERwhere for five bucks a month,

1:05:22SPEAKERyou get access to the Discord,

1:05:24SPEAKERyou get access to the monthly Patreon episode,

1:05:26SPEAKERwhich was uploaded from a, well,

1:05:29SPEAKERlet's say 35 to 22,000 feet this month.

1:05:34SPEAKERApologies again.

1:05:35SPEAKER- It's the real Mile High Club.

1:05:37SPEAKER- Yeah, exactly, hell yeah.

1:05:39SPEAKERDoing it, I'm doing it, doing it at 35,000 feet.

1:05:45SPEAKERWe'll be back next month with another Patreon episode.

1:05:47SPEAKERAnd if you had problems getting access to it,

1:05:49SPEAKERbecause I posted it a day late,

1:05:50SPEAKERplease send me a message either on the Discord

1:05:52SPEAKERor on Patreon and I'll take care, I'll hook you up.

1:05:55SPEAKERI don't think that we had any,

1:05:56SPEAKERI looked and I didn't see anybody that was over

1:05:58SPEAKERin the 10 or 12 hours or whatever it was that it was off,

1:06:01SPEAKERbut please let me know.

1:06:03SPEAKERAnd yeah, as always, patreon.com/techpod

1:06:07SPEAKERfor access to that and everything else.

1:06:11SPEAKERThe Discord has been absolutely hopping lately

1:06:14SPEAKERin the Linux channel for, I'm not exactly sure why,

1:06:17SPEAKERbut a lot of good tips in there.

1:06:19SPEAKERAnd then we had a request for a ham radio channel

1:06:22SPEAKERthe other day.

1:06:23SPEAKER- I can see it.

1:06:24SPEAKER- I mean, yeah, I can see that.

1:06:26SPEAKER- I'm not opposed to it.

1:06:27SPEAKER- People want knobs, we'll see.

1:06:29SPEAKERIf there's enough chatter in the common areas,

1:06:31SPEAKERwe'll add in a channel, that's the rule, so.

1:06:35SPEAKERAs always, thanks to everybody who supports the Patreon,

1:06:37SPEAKERand thanks to everybody who supports the show,

1:06:39SPEAKERbut a very special thank you to our executive producer

1:06:42SPEAKERpatrons, including Jason Lee and Felicitous Rips,

1:06:46SPEAKERAndrew Slosky, Jordan Lippett, Bunny Money,

1:06:49SPEAKERTwinkle Twinkie, David Allen, James Kamik,

1:06:53SPEAKERand Pantheon, makers of the HS3 high-speed 3D printer.

1:06:57SPEAKERThank you all so, so much.

1:06:58SPEAKER- Yes, thank you to all.

1:07:00SPEAKER- That'll do it for us this week.

1:07:01SPEAKERWe'll be back next week with another tech pod.

1:07:03SPEAKERUntil then, please consider the environment

1:07:05SPEAKERbefore printing this podcast.