The following transcript was generated by AI and may contain inaccuracies.
Si: Ladies and gentlemen, boys and girls, and anybody else who happens to be listening, of whatever race, species, or gender you wish to be, welcome to the Forensic Focus podcast. We started—we talk off air before we come on air. And we were just saying that we started this year, this is 2025, we’re doing a wrap-up episode for the end of 2025. And we started this year with an introduction to 2025. We were talking about Donald Trump coming to power and what a rollercoaster this year has been.
I’m gonna say that seems so long ago to me, I can’t even fathom the changes and the things that have happened in the world since we recorded that—that was still something that was in the future.
Desi: Yeah. And you even went through and put together all the episodes and it has felt like a busy year and it felt busy for the episodes. But even now, as I sit here, we haven’t done an individual episode. We barely talked to each other this year.
Si: Yeah.
Desi: I think we tried a couple of times to have an episode with just us and then it was back and forth, us getting sick, which kept canceling. That’s actually caused a few of our episodes this year to—we’ve still got one we’re trying to chase down, but between the three of us, we keep getting sick. It was pretty quiet for everyone this year, but I know for both of us it feels really busy, so I guess it’s just the same for everyone.
So what have we got here? One, two, three, four, five, six, seven. Yeah, only seven that we did. Plus the mental health series, which was probably a load.
Si: I’m rather annoyed because he pinched one of my guests. I had Tom from Semantics 21 lined up to talk to because they have a really interesting product. You’ll like it. I’ve persuaded him. He showed it to me—oh no, I did do another conference this year. I was at the CCTV, the policing, NPCC, the National Police Chief’s Council CCTV conference. Oh no, I’ve done two conferences.
God, they’re just so long ago in the distant past. This is what it is. But anyway, I met him at this CCTV conference and he demoed to me a product that they have. And I hate to—I’m gonna have to eat some of my words in this podcast, which is embarrassing, but I’m starting to like AI. I still think it should be referred to as applied statistics, but I’m starting to see that it actually has value.
He’s demonstrating a product that they have—first of all, it’s all offline, which gets rid of a whole bunch of my reservations about putting data into it. But it triangulates locations of imagery. And it’s really impressive. I said, look, you’ve gotta come onto the podcast and I will give you a selection of random photos. You’ve shown me some things that you do, but it could all be smoke and mirrors, Wizard of Oz stuff—you’ve picked the examples that work.
So you come on, I’ll give you three or four separate images, or we would both give—you could give over a couple that you’ve got, ’cause you’ve traveled extensively as well. And see if they can triangulate those for us. But in the live demo, it was very impressive. But he got poached by the mental health podcast, which was infuriating. He’s gonna come on, hopefully—I saw him at the F3 conference and hopefully he’ll be on later in the year.
But no, there’s probably about the same as us, if not more, six or seven episodes in the mental health series. I only listen to one podcast now. And that’s a Linux one.
Desi: Yeah, same. I used to listen to six or seven. But yeah, this year I’ve only been listening to one. I do some audio books every now and then and I’ve pretty much cut out streaming. I don’t know whether you are—how many streaming services are there now? There’s more streaming services now than we appear on podcast platforms.
Si: That’s true. We’ve cut back. So do you have the concept of a television license in Australia?
Desi: No. You guys in the UK are weird. Why would you need a license for a TV? And I know that you have this and we don’t.
Si: The theory behind it is that it’s a tax which goes towards public service broadcasting. Your TV license pays for the BBC, the British Broadcasting Corporation. You get two channels for that. There are two terrestrial broadcast channels—BBC One and BBC Two. And then there’s a whole bunch of radio and digital stuff.
But if you look at the terms and conditions of the television license, it is only required if you watch BBC One or BBC Two programs, or you are watching any live transmitted events. So if I wanted to watch football, which I don’t, but if I wanted to watch football live, I would have to pay for a TV license. That is the legal obligation for it.
And we looked at our viewing. The BBC, bless its cotton socks, has degraded in quality massively over the last 10 years. To the point, there are no shows that I watch on the BBC. I don’t watch sports. And even if I did wanna watch them, I could watch them the next day and it would still be free if I could find them somewhere.
So we’ve canned our TV license. I’ve had a TV license in the UK for the better part of 25 years, and this year we canceled it. We’ve gotten rid of it.
Desi: How much was it?
Si: It was about £170 a year, so about 15 quid a month. It’s the same as a subscription that gets you two channels.
Desi: Two channels, yeah. So our equivalent of the BBC in terms of size in Australia is the ABC—the Australian Broadcasting Corporation. They do TV and radio, and they put out a lot of good podcasts as well. They’ve got their own app, so it’s all in their own little ecosystem. But that’s all taxpayer funded, government funded. We just don’t see it as a license.
And we pay a shitload in tax in Australia. So that’s probably where our subscription for ABC is somewhere in the monthly.
Si: One of the ideas behind the BBC was that they were publicly funded so that they could be unbiased. And then they probably blew that—same ABC—out the water a few weeks ago by being biased to the extent that Trump is suing them because they were anti-Trump. I’m glad that I’m not paying their legal bills now by not paying the license.
But no, you’re right. I mean, streaming services—we’ve got in the house Disney Plus, Netflix. We don’t have Apple at the moment. But we do have some advertising funded UK ones. We’ve got ITV and Channel Four, both of which are terrestrial broadcast but funded by advertising. You watch a program, you get an ad break, and that’s what pays for it.
Desi: Well, if you don’t have premium Netflix now, it’s ads anyway.
Si: That’s true.
Desi: So, as a segue into topics we actually talk about on this podcast, I wonder if there’s ever been a case where they’ve done forensics on someone’s viewing time on streaming services.
Si: I’m not gonna say too much because it’s an active case. But I am aware—I am involved in a case whereby somebody’s gaming habits are questionable as to whether they have an alibi for the fact that they were online playing a game at a given point in time versus out killing someone.
Desi: Wasn’t there a case that we spoke about when we had a guest—I can’t remember who the guest was or what the case was—but it was someone who was streaming their game, but what they’d done was prerecord it and line it up and then go out and commit the crime?
Si: Yeah, it was a live stream. It was a prerecorded stream.
Desi: Yeah, they were like, “Oh, we were streaming at home, how could we go out and commit the murder?” But then they proved that it was recorded, and then they played it through the live stream.
Si: I forget what that was. Well, I’m gonna say it was either the case or a plot of a Death in Paradise episode that did show on the BBC. Same sort of thing. But I have a recollection of that, and I do also remember very well Professor Sarah Morris when she was on, telling us about the guy who was trying to use his washing machine as an alibi because he’d turned it on. But that was turned on remotely over the internet.
Desi: Yeah, the weird digital forensics tech that you can do. That was interesting. And I guess it’s just getting more and more ingrained. Like there’s been a few cases now with wearables. There was one that we did with [unclear] when he talked about—the person who had been murdered—they figured out time of death, their heart had actually stopped because they had a fitness tracker on. Time of death from an Apple Watch.
I guess it’s just gonna get more and more ingrained as we have more wearables. As batteries get better—I think we talked about this last week, maybe off air—having glasses and a watch or glasses and something else, it’ll be more popular.
Si: I think that’s something—it’s an interesting segue into what we think the future of forensics is gonna hold for us, because we are seeing more wearable devices now. Although strangely—you’ve got a wearable on. You’ve got an Apple Watch. Sorry, you have a watch.
Desi: No, I have a very basic—this is a Samsung product, but it’s the Fit 3. It’s the most basic version you could get. It’s got a heart rate monitor, which is fairly accurate, and it counts steps and it just syncs. It’s Bluetooth sync with my phone. That’s it. It doesn’t have 4G, 5G. It’s not very smart. It’s essentially a pedometer with a—
Si: I’ve been through three smart watches. I was a very early adopter and I had a Sony watch before these things were a thing. Do they even have a wearable department anymore? Oh, Sony—I’ve got one even older than that. Hang on. Funnily enough, I’ve been tidying up so I can see.
These are the oldest ones. This one—I might have to turn my blur off. I might hold it up to the camera.
Desi: It looks like a black square.
Si: It is. It’s a black square. It’s a Sony—that’s a Sony Ericsson thing that used to work with your phone to give alerts and stuff. It’s got a couple of buttons on the corners so you can read text messages and things. Really old school.
That’s a Sony SmartWatch, although “smart” being a relative term. It’s got no sensors on the back, just the charging port. But that was a touch screen as well. And then I’ve had a Garmin. Garmin’s been the one that I’ve probably used the most. That was a good product. That’s the smartest of them. That had NFC and stuff.
But this is the watch I actually wear, which was a birthday present. It’s a Casio G-Shock. So it’s about as indestructible as you can get. But this has Bluetooth on it as well. It doesn’t do anything except I can set alarms on it without having to press the buttons, which is useful.
Desi: I had a G-Shock as well at one point. I think I’ve still got it. The battery’s dead, but it had Bluetooth. I could get vibration notifications. So if I got a message, it would vibrate and I could program the hands—the hour and minute hand would turn to a number and I could pre-configure what that number meant for the notification.
So if it was a Facebook message, it would be one o’clock. If it was a WhatsApp message, it could be three o’clock. So I could set which apps were giving me notifications to which numbers. I could have 12 different notifications as long as I memorized which notifications went to which number.
Si: That’s brilliant. I love that. That’s such a good idea. Bring that back.
This one has two Bluetooth features. One is that it will sync the time with the NTP server, so it’s accurate. And the other thing is there’s a button on here that I can press that will find my phone when I’ve lost it, because it will set the phone ringing. When I drop it down the sofa, it’s useful. But apart from that, it’s pretty pointless.
What I will say though is it’s solar powered, which means I don’t have to worry about the batteries anymore, which is a heavenly thing.
Desi: Yeah. So the Sony SmartWatch—they were released, well, Sony SmartWatch 2 was released in September 2013. So those kinds of smart watches have been on the market for a very long time. It was running Android Wear, which is obsolete. What are we up to now?
Si: Android—this being a Google Pixel—is actually current. Although that’s gonna run out soon. It is a Pixel 7.
Desi: Android 16 was released June 10th, 2025. So Google might be on some beta version, but 16 is probably the standard for a lot of people.
Si: Security update October 5th, 2025. Kernel version Android 14. That makes no sense at all, does it?
Desi: So in terms of predictions—I think the wearables will increase probably in form factor of watches and glasses. What predictions do you have for technology? Consumer technology?
Si: I think what we’re starting to see, given the normal technology curve—I think we are in a bit of a lull. We are seeing incremental improvements in lots of things. Everything has more storage.
I got into a bit of a comment thing on LinkedIn the other day about this. Somebody was saying, “We’ve got five petabytes of data off 3,000 devices.” And it’s just over a terabyte per device. We’re well beyond having one gigabyte of storage on a phone. You’ve got 512 gigabytes on half the stuff. You’ve got terabytes on some phones. In computers, you’re on terabytes as a minimum.
I found that figure to be quite low and I was surprised that it wasn’t a higher number. But things like battery life and data storage and speed are incrementing well. But actual new functionality, new stuff—I don’t think we’re really seeing that much.
The only places we’re seeing the huge changes is in the machine learning AI space, where it’s gone from being a clunky, difficult—not difficult to use, clunky thing that’s obviously AI—to a quite seamless thing to talk to and get answers out of.
Desi: That’s probably it. Not so much in the direct consumer space, but something we’ll interact with in terms of businesses. AI-generated people will be good enough to pretty much fool the average person in terms of interaction.
I’ve seen in New York, they’ve replaced cashier workers in cafes with iPads with virtual assistants from the Philippines to cut down on wages. They can pay a dollar 25 an hour to this Filipino worker just to take your order. And then that gets automated through their system.
The next step on that will be when AI is good enough and you have an avatar. Your interactions are minimal. You’re gonna walk in and think this is someone from the Philippines you’re talking to, but it won’t be—it’ll be AI.
Si: Did you see there was an article the other day? There’s a company that does AI transcription for meetings, and when they set it up, the two guys who set up the company didn’t have any AI technology. All they were doing was connecting into meetings and transcribing it all themselves by hand during the meeting, then signing out and sending this “AI created” thing. Total smoke and mirrors.
The articles were all about everybody’s promising us loads of stuff with AI and it’s all rubbish. And this product isn’t—it doesn’t even have AI behind it.
Desi: I hate that. It happens all the time though. It’s just marketing and advertisement.
Si: There it is. A billion-dollar US AI company co-founder admits that its £100-a-month transcription service was originally two guys surviving on pizza and typing out notes by hand.
Desi: Oh, fireflies.ai was a company as well, which is pretty big. We’ve seen that in a few cases. We have the thing at our work where people can’t use transcription because it’s data loss—it’s being stored on cloud servers outside of a company’s control.
Si: I was in a meeting with the Fireflies agent. And I wish they’d sent people because the transcription service was terrible. It was very poor quality. I had a meeting with the same people today and I saw the bot attempt to join and I canned it because I didn’t think it was worth it and I didn’t like it.
Desi: Yeah. So I think that’ll be interesting in the future as these companies push AI companions or AI assistants into more of the wearable devices. Especially with glasses that have the video panel.
Do you ever play—have you played Zelda? Growing up at all?
Si: Hasn’t everyone played Zelda?
Desi: You know how there’s always that really annoying companion?
Si: This is it. We’ve gone from being competent people ourselves to having Clippy sitting on our shoulder telling us stuff. And remember what happened to Clippy the first time around.
I’m gonna say—for the young, Clippy is probably before your time as well. But at least you’ve been in the IT industry enough. Anybody not listening—look up “Clippy Microsoft.” That was recognizing behavior and then offering to assist on the basis of behavioral identification. “It looks like you’re trying to write a letter. Can I help?” Yeah, no, go away. I know how to write a letter.
We keep trying to do this to ourselves. I think it’s interesting that the human idea of utopia is something you can talk to that does your work for you. It’s less humans.
Desi: We’re all secretly introverts. If there was just less people and I could just talk to someone who wasn’t a person, that’d make my life so much better.
But where I was going with that was—as more of those companions come in, ’cause we’re already seeing it in courts now where AI companies are being asked for data. Sometimes it’s legitimate, sometimes it’s just asking to try and see.
There was the one with OpenAI where the New York Times was suing OpenAI to get access to a bunch of data because they were doing a story on whether customer’s data was de-anonymized or something. And OpenAI was like, “No, that’s a breach of privacy.” And they’re like, “Is it a breach of privacy if you’re de-anonymizing the data?”
It’s this back and forth. “This is part of legal discovery. We’re not releasing the data. We’re trying to legally prove whether or not you’re de-anonymizing it or not.” Anyway, they ended up losing—the injunction they were trying to do to stop it. The judge said no, you need to provide the data.
And then OpenAI put out an article to the general public, like, “New York Times is doing this,” but missed out a bunch of the details, including the fact that they’d lost the injunction and the judge was like, “You need to provide the data.”
I think we’ll see more of that with AI companies. But it’ll be interesting when those companions are more ingrained in our technology and our lives. I assume in the glasses they’ll automatically be filming. Then Clippy will pop up and be like, “Oh, I see you are trying to break into a car. Would you like some help with that?”
And then how the legal system goes about trying to obtain that from the company. It’s not gonna be stored locally, it’ll be stored in the cloud somewhere.
Si: It’s interesting you should say it’s not gonna be stored locally, because some of it is stored locally. From what exists now, there are artifacts that can be found on devices that show queries, responses, images—both pre, during, and post-processing.
The example that I saw was somebody putting a bank card photo into an AI asking it to change the name on it and receiving a result with a changed name. But you had all—you had the first part, you had the second part where it had figured out the parts that needed to be changed and had blanked them. So you could see the parts that were going to be changed and weren’t gonna be changed in the image. And then the final completed image, which put—I think it was John McClane from Die Hard’s name across this bank card.
But circling back slightly—there’s another good lawsuit in the US, which I’m part of. It was a class action against one of the AI companies for hoovering books. And they lost. And the settlement is billions.
Desi: This is the one where they were pirating books and then just ingesting it as part of their learning model.
Si: Anthropic agrees to pay $1.5 billion to settle book piracy lawsuit.
Desi: Yeah. And this is the sad part. They’ve already benefited from ripping off the works and there’s no takebacks. There’s no tearing that out of the learning model. They’re paying what they should have paid for in the first place. But they’re now so big that that payment is worth it. They’ll just pay this settlement. They’re gonna make way more money and they already have all these works now.
It’s shitty that they all did it—Meta did it. I know Meta’s in part of it because they had those email leaks where they’re just like, “Oh, should we be hiding our IP address?” And they’re like, “Because we’re pirating books and we don’t want it to come from Meta.”
I’m sure a lot of the other big ones have done the same. They were just like, “Oh, all this content that we can just ingest and use.” But it wasn’t in good faith, which is shitty for the creators.
Si: Although hopefully I might get a payout for that. I’ll let you know what my share of $1.5 billion is once it’s divided amongst all of the authors, all of the publishing companies…
Desi: And once all the lawyers have taken their fees.
Si: I’ll send you $1.50 when it arrives. But yeah, fascinating.
Somebody pointed it out on LinkedIn—a gentleman called Angus Marshall, a professor in University of York, I think. He pointed it out on LinkedIn and they made it really easy to search whether your book had been included in it and then put in the paperwork and join the class action. It was a very well-managed thing, though they’re gonna take a healthy cut out of it.
But this is the real problem with AI at the moment—the ethical implications. We’ve got stuff that’s completely plagiarized. All of it’s plagiarized to a certain extent because it’s a conglomeration of the works of millions of people across the internet, whether it’s been copyright protected or not. It’s still taking human creativity and using it as input. So you’ve got that ethical dilemma.
And then you’ve got the inverse—the article I sent you about the fact that it’s now being used in the legal system in various ways, and we’re having to filter out where it’s appropriate and where it’s not. And that’s not been a well-defined thing yet. I don’t think anybody’s making huge inroads into it either. Law moves at a pace that makes the average glacier look speedy.
Although, they’re talking—bad news of the day here, and this will date this particular episode—they’ve decided, or certain parts of the government in the UK are trying to, get rid of trial by jury for certain crimes. Given that being tried by a group of your peers is a fundamental basis of quite a lot of our legal system for the last 800 years, all of a sudden to try and rip that out of the legal system is a bit dishonest in my opinion.
Desi: Yeah. The article that Si was talking about was “Judges have become human filters as AI in Australia courts reaches unsustainable phase, Chief Justice says.” It’s saying that lawyers are using AI and judges are having to spend more time being a filter.
I saw another good quote, and I’ll misquote this, but it was: “AI speeds up implementation, but we need to equally slow down our thinking.” It showed an example—some AI consultant had gone into a company. They’d shipped an AI feature in two days, and 13 days later they had to pull it because they’d implemented, hadn’t really thought about what the impact was gonna do to the product. All the legal problems with data. Hadn’t thought of that.
So they had to pull the product and spend the next three months trying to fix that. And I think that’s everywhere.
In the legal system, before AI, lawyers would’ve put it all together. And as a judge, you’re trusting the level of research and rigor that goes into putting something together and presenting it to the court. You use less thought cycles as you’re reading it because you’re like, “Okay, they’ve quoted this case, they’ve obviously gone and looked at the case.” You might go look at it if it didn’t sound right, but most of the time you just trust it.
Whereas now we’re seeing examples of AI just making up court cases that don’t even exist. Just to support its point.
I think that’s everywhere because I was using AI at my job the other day. It was in our corporate one, in one of our code bases. I said, “I need you to go search all of our content for anything that includes these keywords.” It was getting it to do a grep search for me.
It took me five tries to get it to return the full list because it kept truncating it. I said, “Give me the full results.” And it was like, “Oh, okay.” And gave me another truncated list. “Why are you truncating the list?” “Oh, there’s a lot to display.” “I don’t care. Show me everything.”
Finally got it to show me everything. I said, “How many file names did you return?” And it said, “Oh, I’ve found 24 different rule files that have these keywords in it.” And I counted—it was 28. “Why did you tell me there was 24 when there’s 28?” “Oh, you’re right. My mistake.”
You can’t even do basic math. In the end, I was just like, “This would’ve been easier if I just opened a terminal.”
Si: And used grep and then piped it into a file and then wc’d the file.
Desi: I didn’t even need a file. I could have just done grep, give the headers where these keywords are in, and then pipe wc -l. I thought this would be easy. I thought this would be one command where it would work.
Si: It’s this counting thing. It’s so simple. I tried to get it to write a poem with a specific number of letters per word. It didn’t—I wasn’t expecting bloody Shakespeare or anything, not that I’m that big a fan of Shakespeare.
But it was just like, “Here’s your poem, and we’re using the word ‘soy’ for the one with two letters in it.” And you’re like, “You’ve got no clue.” I don’t know why—well, I do know because it has no thought. It’s doing a probabilistic return on what the next most likely word is. That has nothing to do with what it’s actually thinking about.
Especially when it comes to poetry, it’s looking at its corpus of previous poetry and finding likely next words. But it’s missing that thinking step to engage.
I’m gonna say ChatGPT has gotten better. It’s the only one I really play with. I have a paid subscription on ChatGPT. If you—it has two options. One is deep research where you can send it off and if you ask it to bring you back Harvard references, it does. And it makes them clickable links so you can go off and check to make sure it hasn’t fictionalized anything. That’s really useful.
Desi: And the researchers at Harvard haven’t fictionalized the AI research.
Si: Am I being paid to review this academic paper? No, it’s been published, it’s been peer reviewed. I’m sure it’s fine. You gotta stop somewhere.
So I’ve started to find it useful. It’s kind of a better search, in this circumstance. It’s quite good if you say, “Can you go off and look up academic papers that relate to this topic in forensics?” It will trawl papers and it does come back with references. No, they’re not always the best, but at least it’s quicker and easier than doing it yourself. And as long as you’re verifying it.
But this is the mantra of forensics anyway. We don’t trust anyone. Everybody’s lying to us. We have to see it with our very own eyes before we believe it. Which is why generally speaking, we don’t like the push button stuff. If I haven’t verified it in a hex editor to prove it’s there, it doesn’t exist.
So I think in that regard there is room for AI in the space. But again, it’s a difficult one.
It’s an important distinction I’ve been trying to make to some people. In the UK and in the court system, there’s a difference between intelligence gathering and evidence gathering. You can use a tool to find something out and carry out an investigation, but that’s not the same thing as providing solid evidence to support what your assertions that come out of the investigation are.
If you’re like, “Can you show me all of the people who have written about this topic?”—it doesn’t mean that one of them has necessarily committed the crime. But if you use that list of people to narrow down your search suspects from 44 million to 12, that’s fine, that’s acceptable. But you’ve then gotta go and get the evidence that somebody’s actually done something.
I think this is where people are treating AI wrong. They’re treating it as a finished product and it’s a tool. It’s to get you the next step, but it’s not the final step.
Desi: A finished product and a decision maker is what some people are treating it as. I think it was Bill Gates when computers were first commercially introduced—he was like, they should never be made—they’re not designed to make decisions. They’re there to assist, but never to help make a management decision.
And I think AI is the same. The judges—they’re still the human. There’s always gonna have to be the human in the loop. But it’s when it becomes so normalized that the human in the loop’s almost ignored. Or we’re just like, “Oh, it’s done this a thousand times. We don’t need to check at all.”
A good thought experiment is self-driving cars when you’ve got no guardrails around it. The Teslas in the tunnels where they literally can’t go anywhere—it’s just brake—there’s no pedestrian walking across the road thought experiment. That’s a good example where you can let it do its thing because there’s no surprise there.
But if a self-driving car’s on the road and a kid runs out, who’s at fault? The person or the car for not stopping in time? With defined boundaries, we can let it make decisions, but you’re severely restricting it. It’s almost like binary choices.
But yeah, we’ll see it more and more in the new year. I don’t even think that’s a prediction. AI is inevitable at this point. What else do we think is happening?
Si: I think there’s a couple of interesting things. I’m not sure how realistic I’m being about it, but I think the operating system wars are opening up again.
It used to be that Unix was dominant, but that was a long time ago. Microsoft cornered the market and owned 99% of it. Apple owned 0.9%, and everything else had 0.01%.
I think it’s shifting back. Apple is a far more prevalent operating system than it used to be—Mac OS, both in iOS and in the desktop versions. Although it’s interesting to see that iOS and macOS are actually aligning operating system wise. The code base is tying up.
Desi: They’ve been doing that for a while.
Si: Yeah, it’s been moving that way. So much so that iOS and the iPad is now a fully multitasking operating system. You can open windows and move them around and have multiple windows open in the same way as you can on a desktop. It used to be a single app at a time.
Android and iOS are the only competitors in the phone market really. There are one or two others and a couple of feature phones, but fundamentally, Android and iOS—with Android having the vast market share, which is Linux.
We’re seeing the Steam machine coming out, which is Linux. The new Steam Box is Linux. I’m wondering if perhaps, with the Android tablets, with iOS, with that—if Microsoft’s actually starting to lose its grip on the consumer market a bit. A fairly controversial thing to say, I appreciate that.
Desi: So Apple’s—I was just looking up some stats. These are from AI, so I don’t know how much I trust this. Current for computers, not mobile devices—Apple has approximately 8.7 to 10.1% globally based on shipments market share.
The other thing—this is only Western world. If we consider the global market, including places like China, there’s probably more players in the field than just Windows and Apple. I don’t know whether they’ve got different mobile operating systems, right, but not—
Si: I think if you look at China, they have Linux-based operating systems. So does North Korea.
Desi: HarmonyOS is apparently a big operating system in China.
Si: That’s Linux based, I think.
Desi: They have something called Kylin, which is a Linux-based one. And KaiOS—that’s a Linux-based one as well. That’s a smart feature phones one.
They also have AI PC, China’s first homegrown PC operating system that features AI capabilities. Which is really interesting. They would sell so many units, which would affect the global market share if you looked at all those stats.
But with the geopolitical stance of the US, UK, Australia—we’re not gonna be buying Chinese operating systems anytime soon. But we also don’t consider it as market share.
Si: Just on the buying Chinese operating systems front—there was another interesting article the other day. A guy bought a robot hoover that was a cheap Chinese thing, discovered that it was sending all of his house mapping data back to China. Which is not terribly surprising—it’s not like this thing has enough processing power to do smart AI features on board. It clearly needed to offload it onto another processor.
But the thing was, he blocked the outward ports for that data export. And the hoover carried on working to a certain degree—was just wandering around, bumping into things and hoovering.
But a few days after that, it got bricked remotely. They sent a firmware update that killed it.
So ChatGPT’s reckoning—across all connected personal devices worldwide. All Android is about 44-45% of the worldwide market. Windows 27%. iOS 16%. macOS 5%. And then everything else—Linux, ChromeOS, Harmony, KaiOS, et cetera—is standing at 7%.
So effectively Windows, which used to be the absolute dominant market share, is now second to Android quite substantially.
If you look at desktop operating systems, it’s 71% to Windows, 16% to macOS, 4% to desktop Linux, 2% to Chrome, and then 8% “I don’t have a clue”—that could either be things that aren’t reporting properly on the internet or genuinely other operating systems.
In tablets, iOS is slightly over Android at 51 to 49%. Mobile phones though—Android 72%, iOS 27%, and then the others are fractions.
Of course, embedded TVs—usually Linux and Android. Of all web servers, 77% Linux. Of the top 500 supercomputers, all of them run Linux.
And also, a lot of people are in the position now with the Windows 11 rollout, which has canned support for anything that doesn’t have a TPM or doesn’t have enough memory or whatever.
Desi: Didn’t they fix that? I know when they originally rolled out Windows 11, it was bricking a bunch of things.
Si: It doesn’t brick them, it just tells you you can’t upgrade. So if you’ve got an old machine…
And Linux has become easy to get hold of and easy to install and nearly easy to use. Well, easy to use for us, but for your average person, it’s actually become a plausible and usable desktop system. I reckon that Windows might start to see a greater decline.
Desi: The one thing holding back Linux from being more dominant for the average consumer is that you just can’t run all the apps that you can run on Windows. That’s from a developer standpoint. They might make their application work on Windows and Mac and that’s it. They don’t make a Linux version.
I think if the majority of things that you could do on Windows, you could do on a Linux machine, everyone would have Ubuntu. That’s the only reason that stops me from having my daily driver.
Si: What do you actually use on Windows that you would miss on Ubuntu?
Desi: It’s probably been at least 12 months since I looked at this, but things like DaVinci video editing—and I think at the time, some games wouldn’t run on Linux. But that’s gotten better with how much Steam has ported stuff across, probably since they made the Steam Deck, which runs on Linux.
There was a few of my incident response tools that wouldn’t run on Linux, so I still had Windows. There was enough for me at the time to be like, “I couldn’t switch across.” But maybe now.
I think there’s still a gap in the consumer market for the average person. But that’s probably diminishing because—I mean, we’ve got a ChromeOS laptop and it doesn’t run a bunch of stuff because it’s essentially just a browser on a laptop. We use it for the fitness business. It’s not my daily driver, but it’s still good enough for—everything’s in browser.
Si: In fact, funnily enough, I think this is where Microsoft shot themselves in the foot. My youngest daughter needed a laptop and I had a really crappy old knackered MacBook Air kicking around. I mean, so bad that if it’s not plugged in, it doesn’t work—the battery is so shot that it literally needed to stay plugged in to run. It had no battery capacity left.
And I put Linux onto that for her to use for a bit. And the touch pad didn’t work properly, which is another small problem.
But she was working on a project with a couple of universities and they all had an O365 instance. So she was logging into O365, doing her word, doing Excel, doing Teams meetings through the web browser. She didn’t need Windows. It just wasn’t a thing anymore.
You could have a Chromebook—and I don’t know whether they did this to counter Google Chrome, Google Docs and Google Sheets or whatever it’s called. But I hate the Google version of the online stuff with a vengeance, to be honest.
Desi: There’s just some—when you get into it, especially Google Sheets, it’s just missing some of the functions that you have in Excel and you’re just like, “Why isn’t this included?” Like, it’s such a basic feature.
Si: The one that drove me round the bend—Word on O365 to start with—was you couldn’t do footnotes. And I probably write more footnotes than any normal human being should ever do, other than Terry Pratchett, which is a writing god and therefore should be emulated at every available opportunity.
Desi: For our listeners as well, if you don’t know who Terry Pratchett is, go out and buy one of his books and read it. It’s really good.
Si: It’s the best. And I use him for teaching ethics because you must—you must have read—from any of the Sam Vimes books. There’s a set of books which involve a police officer, an officer of the Night Watch in this universe.
Each of the books covers a particular ethical dilemma. You have Thud, which is racism—although it’s speciesism because it’s dwarves versus trolls. You have—is it Raising Steam?—which I think is modern slavery. There’s my personal favorite, one called Night Watch, which is just fundamental ethics of policing as a whole and the rights that a police officer has.
Desi: Because that’s before he is a captain. I think Night Watch is the first one.
Si: The first one is Guards! Guards! There’s a friend of mine called Fran Peacock who’s a CCTV expert in Ireland now. She studied science fiction and literature to a master’s degree level. She’ll correct me if I’m wrong, but I think Guards! Guards! is the first one, and he starts off very lowly and ends up as the Duke of Ankh-Morpork—so very high ranking aristocracy by the time he’s finished.
But yes, they are very good. It’s interesting to see ethics sneak in under the guise of humor. Very important things get said. I do honestly teach it.
Where did we get onto this? Oh, footnotes. That was it.
So I think we may see a dynamic change in the way that people relate to computers in the home. Things like the Steam Deck coming along—not the Steam Deck, the Steam Box, the new one—for gamers, which will replace the traditional console with a computer again. Although if you look too closely at either a PlayStation or an Xbox, you realize that they are just computers anyway. With Nvidia cards and all of the usual goings. Just a scaled-back OS.
Desi: I think because—that’s what I’ve got. In my home, I’ve just built my own Linux PC, stuck Steam on it. You can purchase—they’re just making it productized for the person who doesn’t know how to build a computer or doesn’t want to.
Si: But what’s interesting is in that productized version, there is a desktop. You can click it to give you a full Linux desktop that allows you to surf the web and stuff.
Desi: Well, we’ve had that. I’ve got a Steam Deck as well. You can exit the handheld mode and it’s a full—
Si: I’ve never actually bothered with mine. I understand you can plug it because it’s got HDMI out on the top and USB-C, so you can plug a keyboard and mouse in.
Desi: Yeah, I’ve got a dock. And the dock connects to a wireless keyboard and mouse. So it’s much easier when it’s on the TV. Then you can go into computer mode, which would be good for the bigger games like Civilization or something where you need the mouse rather than a controller.
But even Samsung for a long time—they’ve had DeX, which is you can put your phone into a dock and then it’s a Linux or Android operating system. They don’t push it as much. I figured out it’s still on my phone. I’ve got a Samsung phone. It’s just they don’t push it as a feature anymore.
Si: They’ve got a dedicated customer base who would die and go away if they stopped using it, but they don’t wanna sell it to anybody new. Samsung’s got a lot of interesting features—along with their Knox, the secure section.
Desi: Yeah, the secure one.
Si: That’s cool.
Desi: It is interesting. I think we are switching from—it’s been gradual over time, but even a generational thing. We grew up using desktop computers and then got introduced to phones, whereas younger generations now start using phones and mobile devices.
And I was talking to someone the other day who was my age—36—and they’d not grown up the way that I did having a desktop computer. Because my dad was just into it, so that’s why we always had one. And then at school.
But she grew up and never really had a desktop computer. Her only interaction was mobile phones as her first interaction. And we were talking and she was like, “Oh, I need to do this thing. Can I borrow a laptop?” And I was like, “What do you mean?” And she was like, “Oh, I just can’t—it’s too hard to do on the iPad.”
She had a work computer that was at work, but she needed to do something at home and didn’t have a keyboard. And I was like, “What? Who my age doesn’t own at least some shitty laptop?” But she just didn’t own one. She owned a phone and an iPad. That was it.
I never thought I would meet someone, but I knew that was the case for younger generations.
Si: Admittedly, they’ll all have computers. It’s my daughter who grew up in this household, and you can imagine how many computers there are here.
Desi: Yeah, there’s probably 25.
Si: I’m not gonna deny it. This is before I even get to the garage.
A friend of mine literally just told me off on the phone because I said I’d thrown out an old IBM Unix server. He was like, “No, you don’t throw away old hardware.” And I was like, “Yeah, but I was only using it to stand on to get stuff off tall shelves because it’s just so solidly built.”
But it wasn’t a server, it was a step. It weighed about 30 kilos. Seriously. Steel frame chassis. It was bonkers. The tip looked at me somewhat odd when I took it over there.
But she wasn’t gonna take that crappy MacBook Air to uni. So I said, “Look, what is it that you want? We’ll sort something out.” I said, “Do you want my iPad? I’ve got an iPad Pro. This comes with a keyboard case, you’ve got the pen. You can make notes on it.”
She was like, “No.” She actually wanted a real keyboard. She couldn’t stand the idea of not having a real keyboard. So there is some sort of—it’s definitely a familiarity thing, not an age-related thing.
Desi: Yeah. Well, I think your situation’s unique. There’s less and less of that because she’s grown up around so much technology. The general population doesn’t. Because their parents may have a computer, but the kids don’t really use it that much. They get an iPad for Christmas and it’s that and consoles, isn’t it?
I see that with my friend that doesn’t own a computer. Her daughter is very much like that. She’ll use an iPad. She has one of those shitty school-funded laptops, which she’ll probably never use again once she leaves the school system. And they have a Switch. That’s her exposure to technology—a phone, an iPad, and a Switch.
Si: How do these people get data into machines? I type faster than they write. And the interface to a tablet to either write and get handwriting recognition or type onto it is awful. How are they—are they talking at them?
Desi: No, but they’re not using technology the way that us and our listeners probably use technology. We’re putting in shitloads of information into our computer for video editing or writing copy or whatever else.
The average person’s not doing that. They’re just consuming content and that’s what the device is designed to do. You consume apps, you consume YouTube, you consume streaming services. That’s it. You’re not actually generating.
Si: I see your assertion and I wonder if you are only partially right. I wonder if it’s just that the content that they are generating is now almost universally video. They’re all uploading photos of lunch to Instagram.
Desi: Yeah, so they’re no longer writing long Facebook posts like my aunties used to do and rant about things. They may upload a picture with a small caption.
Si: That’s probably true. And also—it’s interesting, isn’t it? The other thing we’ve noticed going over time is we used to see lots of landscape media and videos. And now most of the media and videos we see are portrait. We’ve gone—
Desi: Yeah, what’s that—nine by sixteen, that Instagram and everything.
Si: Yeah, so I think what we will see in terms of trends perhaps is directed at that. We are looking at things that will enable the creation of media in that way. And it will be things like you suggested—cameras in glasses and the phone that you tap to start and stop recording.
And then we’ll be looking to extract information from those devices. Although at the moment, I’m not sure there’s enough processing power in either the watch or the glasses. There will always be a third phone device in the pocket doing the hard graft.
Desi: I think the Apple Pro Watch probably has enough processing power. It doesn’t—it wouldn’t have enough battery. I think battery is definitely our limitation. If they can solve the small-factor battery problem—because a lot of those devices have smaller processors because they’re more energy efficient.
But if you could have more energy, then I think it would eliminate phones—not eliminate them, but people who were adopting the smart watch, smart glasses form factor probably wouldn’t carry their phone.
Si: I think the one thing that I’ve also noticed—mercifully, not universally—is that people seem to have changed the way that they use their phone.
Those of us that are old enough to remember things that were connected to the wall—that you actually had to hold. The speaker to one ear and the microphone to the other, to talk into it. Still use a telephone in that way.
People wandering around talking into it like this, and all of us hearing their conversations as it goes on—drives me around the bloody bend. And that’s just me. I’m gonna get to a point soon of yelling “Get off my lawn!” at people. Getting to that age.
But I think if we could get rid of the phone so that people can just have bone conductive sound or earbuds—I use earbuds a lot for calls because I don’t particularly like holding a phone.
Desi: It would be super interesting to get someone who is young in digital forensics onto the show to talk to them about how they use technology. Or even just someone younger in general from the more general population.
I was reading an article the other day—it might’ve been Gen Alpha or Gen Z—about how they were essentially, “We came up with this idea where we were just gonna chain our phones to the wall or to a band.” It was making their mobile a landline again.
They’re like, “We use our phones too much and we don’t wanna use them.” And even though the use reduction applications where it strips all your applications down to the bare minimum wasn’t working. So what someone did was create a case that they could lock their phone into and then they chain that to the wall. If they wanted to make a call or anything, they had to get up and go to their phone.
And I was just like, “That’s just a landline.”
There was this—I don’t know how big the movement was—but there were a few people doing this because they’re just like, “We spend way too much time scrolling on our phone. We just wanna embrace—” it’s that whole “embrace boredom” movement. I don’t know whether you’ve seen that.
People film themselves doing it and then post that as content. How is that content? They time-lapse it, so at least it’s faster. But they film themselves for half an hour, an hour, just sitting there doing nothing. And the whole point is to be bored, not to consume much. It’s to try and rewire your brain to be able to not have a device.
Si: Part of me quite likes that as a concept. Yes, the 24/7 media consumption is probably not great. The other part of me loves the irony of videoing it for content purposes. Why—just go outside for God’s sake.
I know you—I’ve seen a few of your posts relating to health, more where you’re saying, “Go for a walk. You’ll feel way better afterwards.”
Desi: And I get this whole content thing. I sometimes—because I use Instagram for the fitness stuff to post—I’ll go on there to post something and then you just get sucked in to the scrolling. It is addictive and it’s hard. It’s like, “Nope, I’m doing work, I have to stop this and do something else.”
Si: Have you noticed a massive uptick in AI-generated material on Instagram though?
Desi: Oh, I fucking hate it. It’s actually helped me get off social media faster because I see it and I’m like, “This is rubbish.” And I get off. And especially on X—old Twitter for those who knew it.
Si: The only time I go onto X is if I am looking to find out what the general British public reaction is to something. And then I just canvas the morons and then come back out again.
Desi: So on X, there is so much AI-generated content. It does a pretty good job at marking what is AI generated—it has a little warning underneath it. But for either really fresh content that’s gone up and it hasn’t got around to looking at it yet, or some of it is fooling whatever mechanism they have to detect it.
There is a disturbing trend for human tragedy to be posed as a news video, and it’s AI content. And that’s deeply disturbing where we’re going with that.
It’s not just like funny—Instagram, I see a lot of AI-generated content that’s just meant to be funny. Sometimes at people’s expense, sometimes it’s just funny cat videos that someone’s AI-generated and they’re making the cat do funny stuff. I think that’s because that’s the platform that Instagram is, and the people generating that content are more lighthearted.
Whereas X, you get more of the extremism content from different political views and also people who are just probably not socially right. And they’re posting this content and they’re thinking it’s funny, and I really dislike that.
And then when it’s not being caught, you see some comments from people where it seems like they’re believing it. And I kind of want to think that maybe they’re just trolls and they know that it’s fake and they’re commenting that, but I’m sure there’s some people that are getting caught up in this content and they think it’s real. And they’re worried about it.
Si: One of the reasons I left X was because there were people who were believing things that no matter what logical explanation was put to them, would continue down that line. And once moderation had disappeared, it just got worse and worse.
I’ve held onto my account so nobody can impersonate me because that would be detrimental. But effectively, it’s not anything for me anymore.
And I haven’t found an alternative to it. I know people jumped ship to Mastodon and—
Desi: Blue Sky’s another one. Blue Sky was Mark Cuban. I believe I have a Blue Sky account.
Si: I certainly have a Mastodon account, but I just never found them to be—I just kind of lost the will to invest in them.
Desi: It was just—Twitter was just a social phenomenon at the time. And it built such a following, it’s hard to create that for everyone somewhere else. Probably the same as what 4chan was for some people, and then it turned into Reddit. Some people probably didn’t come across to Reddit.
But the amount of AI-generated garbage and just disturbing—what’s worse is some of the stuff that I see that I know is AI-generated content—it’s getting pretty good. It is really hard to tell in some of these.
Si: And the problem is it’s harmless and you look at it and you think, “Ah, it’s just a picture of a raccoon. It’s cute.” But the reality is one, it inures us to seeing AI-generated content in general and we start to accept it as more acceptable.
Desi: Well, I think the danger here is—on X, the stuff that I’m seeing that’s more extremist. When you take somewhere like the US where people will violently riot over—or there is a clash from peaceful protests and then police over something that happens, particularly something around race.
If there is a piece of AI-generated content that’s viral that causes something like that, then something that never happened is causing a real-world consequence. And I think there is a real risk of that.
Si: I think there is a risk of it in the UK. The impression of Australia is that you’re all way more laid back than we are. But I think we’ve seen that—we’ve seen a certain extent here with immigration and issues around housing, asylum seekers and stuff like this.
I think it’s awful. These are people who are coming to our country for help, and treating them in such a way is completely despicable and certainly isn’t a thing that I would wish to be associated with or for my country to be associated with. But it is what it is.
But even that which got quite violent—not every citizen is carrying a machine gun, which the Americans are entitled to do. And therefore it does seem like a really bad idea to rile up a country which is able to riot and is also armed.
For all our faults and our ability to get the wrong end of the stick and go and camp outside a hotel with the asylum seekers, at least not everyone’s packing.
Desi: Yeah. So I think X is all about free speech, but there needs to be some accountability for—especially for generated content. I don’t think AI-generated content should fall under a free speech category. And I don’t know from a legal standpoint what that means.
Si: It’s interesting, isn’t it? If I paint a picture which is offensive, I can claim it’s art. If I tell an AI to create a video which is offensive, is that art still, or is that just offensive? Philosophical questions—I don’t expect you to have the answers to.
Desi: Well, I think this is interesting because we’re trying to make the argument that an offensive piece of art, if that’s allowed because it’s protected under artwork, then we should allow it for AI.
I think the difference is scope and scale of art. If you make an offensive piece of art, you can say, “Sure, this is offensive. We won’t publish any pictures of it. We’re gonna put it in a gallery. If you’re offended, don’t go to the gallery.”
Whereas if you generate a piece of AI content and there’s no moderation on a platform, you are part of that social platform and that’s ingrained in our everyday life. You then no longer have a choice whether or not you’re seeing it.
And the same argument could be, “Well, don’t go on the platform.” But the platforms don’t want you to do that. They want you to stay on the platform because that’s how they make their money.
So I think there needs to be a level of moderation if you are exposing people to content that they don’t wanna see or they find offensive.
Si: Yeah, I think that’s reasonable, isn’t it? You see some warning of it. You see it some of the time at least for things like “this is over 18” or “not safe for work.” You have the moderation guidelines on it that inform you before you start consumption of that media.
So perhaps it is that—but it comes down to the detection of AI then, or is it just simple, straightforward moderation? Are we just gonna pull out anything?
Desi: Well, there’s two things there. One is tagging whether it’s AI or not. And this is where I don’t know where the legality of this is. But if you’re saying that AI-generated content is not free speech, then it’s tagging AI and then moderating anything that’s AI that’s offensive and is not protected by free speech.
Si: Does Australia have the concept of free speech? Ours—the UK concept of free speech—is not as free as you might think. And the American one is extreme.
Desi: We don’t have the same as the American one. So the last time I looked into this—you can say whatever you like, and as long as you can’t prove that you directly incited violence in the US, you can’t be charged for that violence.
Whereas in Australia, we have looser guidelines. So you could say something and still indirectly cause something and still get charged for it in Australia. So our free speech isn’t as free because it’s like, “You need to be civil and not cause unrest and violence and that kind of thing.”
Which I think is similar in the UK, right?
Si: It is. There are some direct prohibitions against religious hatred and things like that.
But also, we have a legal system which has some quite interesting definitions. For example, assault as a charge doesn’t necessarily involve you physically touching somebody. You can yell something offensive at them, and if it causes them to feel afraid, that’s assault. So you can charge someone with assault for making them feel fear.
Desi: I think that’s the same over here.
Si: So you are not necessarily being charged with breach of a freedom of speech kind of law, but you can still be locked up for being an asshole. Which is a good thing.
Desi: Yeah. It is interesting. And I’m excited in some ways, because objectively, we’re still living in the best time in history to live.
Si: Definitely. As soon as you need medical attention, it becomes very clear that you’re living in the best time in history.
Desi: Yeah. Medical technology and advances. Generally, for us very privileged people, living in a society where fresh water is not an issue to get, food availability is pretty good.
But then at the same time, feeling like you’re on the edge of a dystopian future with technology is the other side of the coin.
Si: Assuming we’ve survived that long, it’d be great.
Should we call it quits roughly at the hour? This is—we are, oh God, we are a month today from Christmas. So by the time this gets edited and goes out, it might be January.
Desi: Because we had the other episode in front of it, I think, that we did with Heather and Jared. So that should come out first.
Si: And also, people—you’re gonna have plenty of time to listen to this because you’ll all be at home doing nothing and just surviving off roast turkey leftovers. So this may be an interesting break for you.
Desi: If you wanna do the embrace boredom challenge, put this on for an hour and a half and do nothing else.
Si: But don’t forget to video yourself doing it and post it too. But with the hashtag #ForensicFocusPodcast, please.
Excellent. It’s been a year—will have been a year by the time people listen to this. And I think we will be back together next year. There’s no intent to not be back together next year.
Desi: Yeah, we will be back together. I may be overseas somewhere halfway through. So we’ll see. If that pans out the way it should, I think we should—
Si: I think we should meet up and do a live show together sitting in front of the same camera, sipping margaritas on the beach ideally.
Desi: We could also do more live conference ones where we go to a conference, we find someone to talk to, and then we can record it there.
Si: First, I’m interested to give this a go. I’ve been threatening to do it for a while and I haven’t actually got round to setting up the equipment for it. But I think at the Forensic Expo this year, I’ll set up a booth at the stand and we can pull some people in.
Now, just as an interesting note—we had a huge amount of success at Forensic Expo this year because somebody came over to the stand. They were asking about careers. I said, “Why don’t you go onto the Forensic Focus jobs site—the job board.” They did, and I saw them at the F3 conference and they’d gotten a job.
So it works, ladies and gentlemen. You can get jobs through Forensic Focus job board. And he was very happy about it, and it was wonderful news.
So I won’t name you because I didn’t ask your permission to do so, but you know who I’m talking about. And I know you listened, so—well, you may have turned off by now in this particular episode.
Desi: The boredom was too much. But we’ll post the link—that’s on our website.
Si: This is the job support on the website.
Desi: Yeah. Which is also where you can find our podcast recordings or the old ones, or on YouTube, or wherever you get your podcast from. If you just wanna listen and not look at Si and the mess that is behind the both of us—blurred tastefully in my case at least.
But we have enjoyed recording podcasts this year. It’s always a pleasure catching up with you, Si, and talking. And I love all the guests that we get to talk to, because I don’t live and breathe digital forensics, but I get to vicariously live through all of you here.
And we’ll be back next year. Hopefully—it was a quiet year for us, as we said—but hopefully with a few more guests next year. And I’m sure we’ll have people like Heather and Jared maybe again. Rob Fried. Sarah Morris—I would love to have her on again and talk to her.
And we have—who’s the one that we keep missing? Oh, Simon Franc.
Si: Yes.
Desi: Yeah, Simon Franc. We will have Simon Franc next year. We have tried to organize that podcast four times. But we will get him on. And lots more of our friends and hopefully some new people that we get to chat to as well.
Si: Yeah, really looking forward to it.
So for those of you that are listening—happy holidays, because by the time you get this, you should be in it, whether that be Hanukkah or New Year or Christmas or whatever you choose to celebrate in your own way. Be with family and enjoy. That’s all it’s about anyway.
Have a fabulous time, and we will catch up with you in 2026. That sounds really weird to be saying it towards the end of November, but forward planning and all that.
Desi: Can’t believe it’s November. All right, thanks everyone. We’ll catch you in the new year.
