 Daily Tech News show is made possible by its listeners. Thanks to all of you, including Tim Ashman, Johnny Hernandez, and Hi-Tech Oki. Going up on DTNS, bad news for vertical farms, why chat GPT caused Google to go into a panic, and how Mastodon could survive its success. This is the Daily Tech News for Thursday, December 22nd, 2022 at Los Angeles. I'm Tom Merritt. And from Studio Redwood, I'm Sarah Lane. From Columbus, Ohio, I'm Rob Dunwood. The show's producer, Roger Chen. My friends, we are heading towards the holiday break. Only two more live shows of DTNS left before we get into our holiday specials. But the tech news hasn't stopped. We'll tell you why in the quick hits. Game on, everyone. Google and the NFL announced that YouTube will offer the NFL Sunday Ticket subscription for out-of-market NFL games, starting with the 2023 season. Customers can purchase Sunday Ticket through YouTube TV or Alucard through YouTube Prime Time Channels. The Wall Street Journal sources say that YouTube will pay $2 billion per season as part of a seven-year agreement. Cora launched a beta for Platform for Open Exploration. That's PFO, but they're just calling it PO, where people can ask questions and have a back-and-forth dialogue with chat bots. PO will access several models, including chat GPT, and will introduce a system for organizations to submit their own models in the near future. PO currently has a wait list for access. And then insert never more joke. The memory chip maker Micron announced it will cut 10% of its workforce. This comes as it's reported it lost four cents per share on revenue that was down 47% on the year in its fiscal Q1 to $4.09 billion. Both missed analyst estimates, which is probably why we're cutting. Last month, Micron announced it would cut chip production by 20% and now expects to reduce its capital budget for new plants by 42% in the year to $7 billion. FTX co-founder Sam Bankman Freed, often referred to as SBF, was extradited from jail in the Bahamas Wednesday and has arrived in New York City. He is charged with eight criminal counts, including fraud, conspiracy, and money laundering. In related cases, the U.S. Securities and Exchange Commission and the Commodity Futures Trading Commission filed complaints against Alameda Research's ex-CEO Caroline Ellison and FTX's ex-CTO Gary Wang for their roles in alleged fraud that used FTX customer deposits for risky investments made by Alameda. Wang and Ellison both pleaded guilty to related federal charges of multiple counts of fraud themselves. Meta's latest update to the Quest 2 headset boosts the maximum GPU frequency 7% up from 490 MHz to 529 MHz. The headset will automatically increase the frequency if it detects that an app can benefit from it. That results in better image quality when using dynamic foveation, for example. The boost works with current titles, but developers can update their apps to boost resolution as well. Sorry, do you think SBF tried to put up his bail in crypto? My first question is, can you bail yourself out of jail? Apparently so. Second is, can you bail yourself out with crypto? I have some FTT, now they're going to be worth a lot if I win my case. Alright, let's talk about food. We've covered vertical farms before on the show as a possible advance in food production, especially for urban areas. Wired's Matt Reynolds has an interesting article out called, Vertical Farming has found its fatal flaw. Europe's energy crisis is forcing companies to switch strategies or close down. The industry's future hangs at the balance. The short version is energy prices rose in Europe by 58%. What was once a very manageable 25% of operating costs for vertical farms, sprayers, LEDs and such became almost half. A secondary issue is that vertical farms are expensive to bill compared to traditional farms and investors are lost tolerant to long lead times to profitability than what they used to be. Yes, you might say, well, all right, is there a solution? The solution could be to charge more, but inflation is making consumers less tolerant of paying extra for produce, especially leafy greens that vertical farms do really well at. With energy costs rising, vertical farms advantages are less pesticide use, less water use. That might mean that vertical farms best bet is to head to arid countries, perhaps in the Middle East, where they also have stable electricity prices. Yeah, I think this is an interesting example of how we take for granted certain things that make it possible for innovations to happen like affordable energy. It's one conversation to have about emissions and all of that, and it's an important conversation to have. Don't get me wrong. But the price of energy being low has allowed us to try a bunch of things. Very similar to how battery technology is a break on what we can do with technology right now that I don't think anybody really thinks about because there just isn't an alternative to the lithium ion battery right now that's worth anything when energy prices go up. Suddenly, something like vertical farms, which has a lot of advantages and a lot of things it can do for certain areas, isn't as possible because now the profitability and the price point have all changed. Yeah, I think that one of the things here that's a key point is that these farms are expensive and more difficult to build. So I think like with regular farming prices go up, you kind of stop doing a certain thing when they come back down, you restart because it's not necessarily cost prohibitive to actually get started in regular farming. But because the vertical farms actually cost more to produce and start, that could be a reason to where this inflationary hike that we're having right now might just say, we're going to move elsewhere to try this technology out. Yeah, it's sort of disheartening to hear a story like this because you go, oh, vertical farms, how cool. You can farm in urban areas that makes it closer to the people who might eventually buy those leafy greens. It makes it possible to grow things that would otherwise need lots of land and lots of water. Lots of land, lots of water. Yeah, like typical farming stuff. I mean, I know it's not the same as putting solar panels on your house, but it kind of reminds me of that, of people saying, well, but yeah, I get it. I get why it's so great, but there was a point, and we've kind of come past that at this point, but there was a point where people were like, but I'm not going to recoup my costs really, besides just sort of doing the right thing. And so I think that we kind of get into a conversation like that with the vertical farming industry is like, doing the right thing, doing the thing that is going to taste better and is just healthier for everybody. If it costs too much, is it going to stick? Well, and I wonder what would have happened if this energy crisis had happened later in vertical farms evolution, assuming it caught on, right? If the sunk costs were in there, you might have a different story going on, but right now it's like what you were saying. It's easier to just adapt in a traditional farming situation, whereas vertical farms just they haven't built up their infrastructure enough. They haven't gotten past that part where they can say like, okay, even we can absorb a little energy cost because we've already paid for the infrastructure. They're still building out. And I think that's, you know, hopefully that won't last forever and won't keep these out forever. And I think the Middle East is a great place for them to incubate for a while until maybe energy prices come back down. You might be saying, but when Sarah brought up solar panels, like, yeah, why don't they just put solar panels on the vertical farms? The fact is that also adds to the cost, right? The upfront cost, right? It would pay for itself over time. It's not the pay for itself over time. That's the problem. It's the like, yeah, but I don't know that we can afford to build that now. Yeah, can you can you make $2 and 10 years or can I make $1.50 next month? Yeah, that's what it comes down to. Yeah. And the farmers around the place that I grew up are like, yeah, I can give you $1.50 next month. I don't have to wait for my vertical farm to pay off. So there you go. New York Times reports that Google issued a code red. Apparently, that's a buzzword in the enterprise these days, skittin' all code red and things. Not in the same way as the movie, I hope. This kind of code red is like, get on board. Let's all figure out how to solve this problem. And Google issued theirs over the rise of chatGPT, Google CEO Sundar Pichai directed teams in Google's Research Trust and Safety Division to assist in developing AI prototypes and products, just bringing more people into the process. Folks inside now, Google, have been debating whether chatGPT could get good enough to replace Google. Instead of searching through a list of results, you just ask chatGPT for what you want and get it. So right now at the moment, the answer is no, since chatGPT still has a hard time distinguishing facts from fiction, but it's going to get better. And at some point, it will be as good as Google search results, which also sometimes surface incorrect information. Google has its own version of chatGPT called LAMDA. There was one employee that was convinced that this was sentient. I remember that from a few months ago. But Google has held it back from public use because of its failings. What do we think? Should Google make search a good version of chatGPT before chatGPT becomes a best version of search? That is a tongue twister of a sentence right there. Yeah, yeah. I mean, but it's a tongue twister of a problem, too. Yeah, it's a good question. I would say the answer is yes. I have this saying, when things change, change things. We know that Google didn't go out with LAMDA. They were worried about what if we are racist? What if we are giving wrong information? There are reasons why they didn't do what they do. But chatGPT has shown that this is really, really some next level technology. So Google, they need to get on with it. Like they called this thing a code red that we need to go and look at this now. Because I do see, when I play with chatGPT, it is the closest thing to the bridge computer from Star Trek that I've seen. I think for a lot of folks of AI assistance, that's kind of what you want. You literally can just have a conversation with AI and it gives you the answers in just a conversation away. If Google is not thinking about that, and I believe that they are, they need to be, because that is where search is going ultimately, in my opinion. Yeah, what I imagine is going on behind the scenes at Google, and I don't work there, so I don't know. But I can see a lot of folks in a conference room somewhere saying, let's be conservative about this. Let's make sure that this tool is ripe for the masses, which it isn't. Our tool, Lambda, in that case. And let's make sure that we're not just unleashing something that's going to break or otherwise going to be misused. And then you've got chatGPT that no one can stop talking about. And all the Google people say, no, come on, we've been doing this for years. We are also doing this. And especially people talking about future of search. Okay, this might be where we're going. And there's no question that Google has already thought about this. They're just not the ones getting all the headlines. I mean, this reminds me in some ways of Netflix saying we want to become HBO before HBO becomes us, right? Google realizing, well, crap, we need Lambda to get good enough to be in our search engine before chatGPT gets good enough to that people stop using Google search, which would be devastating. That's where all of Google's money comes from. It's from ads on search. So they not only need to get Lambda good enough, they also need to figure out how to monetize Lambda because that's a different situation than a list of search results. So when I start thinking about it that way, it reminds me of IBM looking at this new Apple computer that people are buying and realizing crap, we need to have a PC, hurry up and figure that out and get it out there. I mean, it just goes to show that Google is no longer and hasn't been for a long time a new company. It is, it is an old, slow company. And this newer startup, OpenAI, has passed it in something that I think Google thought, well, we have really good AI chops, which they do, but they move too cautiously, they move too slowly than the smaller competitor could hear. That doesn't mean I think chatGPT is ready to replace search tomorrow, but it's a race now to see who can get it good enough to do that. And I don't know what the timeline is, but I fully think that you're going to get your wish, Rob, that you're going to have that Star Trek-like experience sooner or later. I wonder if Google is looking at this and they're serious enough about it that they may make an acquisition. If Lambda isn't where it needs to be, do they just try to go out and acquire chatGPT? I haven't heard any rumors on that, I haven't heard any news, but it's something that kind of is just popping in. ChatGPT is owned by OpenAI, which is funded by one Elon Musk, so... Yeah, it's unlikely, but I see where you're going with this. Yeah, could they buy somebody else who's got some chops that they could throw in there? That's the... When you're a company the size of Google, it's like, what do we have? Google, Meta, Amazon? I mean, they're Apple. There's a few companies in that category where it's like, everybody's suspicious. Hmm, you got your own tool. Do you Google? Well, you know, is it as good as chatGPT? OpenAI made it, not Google. If Google is the parent company of a tool like this, well, you get a lot of finger-wagging from folks that... I mean, there is that anyway, just about the technology in general. But I think it's a blessing and a curse to be a company that is as powerful as Google is to be able to say we are and possibly have the better version of this technology, but we have to be so careful about releasing it because anything goes wrong and it turns into Google's evil. Also, keep an eye out on pedals. We talked about pedals earlier on DTNS. It's backed by a hugging faces, big science project. It's open source and distributed. It's that horse that's in third place right now, among the three we're talking about. There's more in the race, but let's just simplify things, right? And it's coming out and on from the outside. It's not impossible that pedals develop something that passes chatGPT and Lambda because it's even more nimble and creates that ability to use the web openly because they have the least reservations on research. Whereas OpenAI, it's doing some things publicly, but keeping a lot of stuff private and Google's keeping almost everything private. This is an interesting race to watch. When people ask you what are you excited about in the future of technology, this is one of those things that you can say, you know what, those chatbots and how they're going to improve search and who's going to get there first. I think that's a fun one to watch. Yeah, I'm definitely looking at this because I've always dreamed of being able to hit something on my chest or hold up a tricorder and just talk to a computer and it gives me all the information of the world that I ask it. So if Google, if they're thinking about this and I understand the issues that they would have with we can't be wrong because we're Google, but they need to start to skunk works up and they need to get going and just explain to earth that, hey, this is experimental technology. It's not going to be perfect. You still go to Google to do your search. This thing is under alphabet is something different until they actually decide it's good enough to bring it into the fold. Rock in a hard place, though, right? Because if pedals is bad, people go, oh, weird open source project. If chatGPT is bad, they're like, no, who's open AI? If Google's bad, blowback, instant condensation and blowback. I think if they were, well, I was about to say if they were smart, a lot of people at Google are very smart, but if you could somehow make Lambda or whatever it becomes in the future part of Google's assistant, then you're getting the idea, right? And you're getting somewhere because you're not replacing search. You're saying you have lots of options and we win 100% of the time. And it's already in there. Google is putting the technology from Lambda and other things into things. It's just putting them in slower than this race is going suddenly. Well, folks, as we get towards the end of the year, we have a lot of last things of the year to check out and you don't want to miss them. If you're a patron, be sure to check out Rogers last column of the year. He shares his thoughts on a comic series illustrated entirely by image generation, by an text to image generator. It's cool. Find it at patreon.com slash DTNS. Well, Mastodon, having a bit of a moment, monthly active users have jumped from 300,000 in October to 2.5 million in November. A lot of people signed up for Mastodon in a month. That's still really small compared to Twitter's 350 plus million users, but it is a huge spike, an 800% rise in a very short period of time. This is a unique challenge though, since no one organization actually runs Mastodon. It's an implementation of the open source activity pub protocol. That implementation is administered by a German nonprofit at join Mastodon.org. Most servers are run on a voluntary basis. If they take money at all, it's usually through PayPal donations or a Patreon, but in almost all cases, you don't need to even pay a dollar to use it. Yeah, each instance's admin is responsible for everything, the maintenance, the moderation, the legal compliance, so things like Digital Millennium Copyright Act, GDPR, COPPA, any local laws in each market where you're made available, which if you're on the Internet, you're in all the markets. Not all of the work is about laws though. Each instance has a community that the admins are responsible for moderating. And all the drama that comes with moderating humans comes with doing it. Smaller the community, the easier it is to manage, but as we said, it's getting bigger. And of course, there's all the maintenance, the uptime, the storage, all of that gets more difficult and grows as your server gets more popular. All of these things take time, effort, and resources to deal with. They're dealable, you just have to increase the amount of time, effort, and resources you apply to them, and that usually means money. Mastodon instances, by and large, do not take advertising. They rely on volunteers and donations, and that can be difficult to scale. Yeah, so bigger organizations are joining in, which could take some pressure off of those smaller instances. For example, Mozilla announced it will launch Mozilla.social, starting in 2023. Tumblr said it would add interoperability with Activity Pub, that's the protocol used by Mastodon servers. The Guardian noted a crypto startup called Social Coop, which runs a Japanese-based Mastodon instance called PaWoo, P-A-W-O-O, which has 800,000 users along with two other instances. And in fact, TechDirt might Mastodon thinks that Mastodon might be about to have its Gmail moment, which is high praise, that moment when Webmail went from being just kind of slow and limited, and it only does so much to the way most people just do email and think about email. So what do we think here? Should we forget about whether it should be a Twitter successor and think of it as something totally different? And if it's mostly volunteer run, can Mastodon actually scale? And how? Rob, you have some experience with scale, I think, right? So, you know, Mastodon essentially is a federated service. And federated services are hard because you have multiple people, and in the case of Mastodon, multiple people who generally aren't paid, who are doing all of the upkeep and maintenance and all of that work. I think that one of the reasons that Mastodon is, and the main reason Mastodon has seen the growth that it is is because of all the stuff that is happening at Twitter. So there's a couple of things that I think could happen. What I think is likely is that it's going to be one of these larger organizations that has the funds to actually stand up an instance and just actually pay people to keep that thing up and running. And you may still see Uber growth there to where it's just going to grow, what is it, 800% over a month and a half. I think that that's ultimately what it's going to take. I told a story in the pre-show about how I logged out of Mastodon on accident. Couldn't remember what instance I was logged into. Had to text somebody to find out, hey, what's my Mastodon address so they could tell it back to me based off of posts that they could see and let me know how to log back into my own account. That isn't scalable at the rate they're growing. Just all these federated services that are out there. But I do believe that if someone with actual money behind them comes in and says, hey, we're going to productize this. We may make it ad supported or just do something to where it's one giant massive server that could literally have hundreds of thousands if not millions of users on that one instance. I think that might actually be the thing that maybe pushes Mastodon over because they've got to react and capture the Twitter people who are leaving Twitter or who are just looking for something other than Twitter relatively soon. The Twitter thing I don't think is going to be a thing for a long time. I think eventually Twitter will kind of settle back into what Twitter has always been. And if they've missed that moment, then you might go back to see the growth go back to what it was like before the Twitter fiasco started to happen. Yeah, I think the challenge with Mastodon, let me start here. The advantage of Mastodon, one of the big advantages of Mastodon is because you can choose different servers, you can choose different flavors of experience. So if you want a more like free wheeling, let people post whatever they want, you can find a server that allows that. Or if you want something that's more locked down, keep all the bad words off, you can do that. If it's fine that Mozilla, Mozilla is non-profit, fine. If Mozilla comes in and makes a real popular server, the Mastodon administration itself has a very popular server with hundreds of thousands of users. Actually, okay, Tumblr comes in. Matt Mullenwig is a good guy. We trust him, right? That's going to be okay. But at what point does somebody come in and make the big popular Mastodon server that then just ends up being the server everybody's on and this whole thing is centralized and the smaller servers fall by the wayside? I think one of the other big advantages is that you could also start your own server. Right now, that's not for everyone because it's complicated enough. But what if instead of just being the big popular server automatic, the folks who make WordPress create the WordPress of servers so that it would be easy for everyone to just roll their own instance and moderate it? Again, might not be for everybody, but you could provide tools that give you the legal shielding, that give you all the compliance issues, the maintenance stuff, the uptime, maybe even moderation tools could be part of that. You get a few of those, a WordPress Squarespace, a Wix of Mastodon servers, and then suddenly you've got a good ecosystem. There could be a business that doesn't have to be fully ad supported in order to survive that keeps Mastodon flourishing and also preserve some of those benefits of being decentralized and allowing smaller niche servers to survive. I agree with everything you said there, Tom. I really think that it's going to take that for them to just maintain this type of growth. I just think it's untenable to have folks who are running these things as a service. This is largely unpaid work. I think that's just untenable at the growth that they're having because things are just going to happen because someone had to go to work. Someone had to, they were on vacation and while they were on vacation, have it happen. I think you're right. I think it could be a combination of the two to where you're going to have some company or some organization come in, stand up a massive server and that allows some of the smaller servers to still stay and do their own thing and be kind of a niche within the Mastodon world. I like Nick with a C calls it the Samsung to Mastodon's Android or maybe to the protocol, to the activity pub is the Android of the situation. That's actually a good good analogy. Yeah, it's a good one. I like that. Well, y'all might be surprised by this, but Nintendo's Game Boy is 33 years old, still has run for president and it maybe it should still has a devoted fan base. I'm one of them. I don't have the original, but I do have a Game Boy somewhere behind me in Studio Redwood. One of those fans though is Sebastian Stax who's created a plug and play device that lets gameplay from the handheld stream to a computer using USB and also be recorded without having to modify the console itself. Now you might say, that's amazing. How did Sebastian do it? Sebastian is a tanker earlier this year. I created a custom cartridge that let the Game Boy stream video wirelessly, which Stax showed off by playing Star Wars on the screen and also a hack that made the Game Boy capable of playing modern games such as Grand Theft Auto 5. I like that. And Mike says, are we still calling it Game Boy at this point? Shouldn't it be Game Man or Game Person? Yeah, Game Grown up. It's a whole adults 33 game adult. Anyway, not to take away from Sebastian Stax, who did a thing that not only will be fun just because you can do it, but Roger pointed out when we were talking about this before the show that you're good for Twitch streamers. Do you want to stream some Game Boy? That's cool. It is cool. Yeah, my first question is like, but why? And all of you were like, because you can. Why not? That's the fun of it. So it's a good reminder that, you know, there are things that are capable of being done and Sebastian Stax has given us a little fun thing to do over the holidays. Thank you, Sebastian Stax for your service. Indeed. All right, let's check out the mailbag. Let's do it. Mark wrote in with thoughts on our discussion we had yesterday about that facial recognition and Madison Square Garden situation with the lawyer being basically identified and thrown out. Mark says, policy and technology are not at odds here. They both function independently for good or bad. The issue with me is that once the technology is in place and they state it's for security and safety, how far do they stretch the definition to include we recognize you and don't want you here when it has nothing to do with the stated purpose of the tech? Mark says, I think the safety and security definition would include disruptive behavior at the venue, concealed weapons, something like that. Is the tech appropriate for that kind of venue? That's another discussion. Mark says if they spotted someone famous, could they draw attention to that fact? Perhaps by bragging about who attended the show. What about using the tech to reward visitors somehow? Is that within the purview of their stated usage? I think not. Rob, you weren't with us yesterday when we were talking about this. What's your take? So I watch everything facial recognition very, very closely because there are ginormous issues with it. This was not a facial recognition story. Someone was recognized. It was by facial recognition. The story is about MSG's rules, not about the technology that they used. Mark asks some pretty good questions. This technology is there. They're using it. What else can it be used for? Fat Joe is at MSG or Jay Lowe is at MSG. Are you going to use it to recognize her and then put the camera on her in the state? They do that already, but can the facial recognition make it easier to find celebrities? That is an interesting question on how this would work. Those are fair questions. What would happen if they're not to me anything about what did happen, though? Nothing you shouldn't be asking questions because that's how we make sure that abuses don't happen, is to talk about, okay, how do we guard against that? I don't know that this particular story is the one that is the smoking gun. I don't think so at all, because if it was a security guard that recognized the lawyer, the same thing would have happened. Well, thank you. Thank you, and thank you, Mark, for writing in. If you have thoughts on anything we talk about on the show, anything we might talk about on a future show, do send them our way. We love feedback. Feedback at DailyTechNewShow.com is where to send that email. Thanks to you, Rob Dunwood, for being with us today. Let folks know where they can keep up with you. Oh, everybody can find me at the SMR podcast, and that is smrpodcast.com, or over at my other podcast, thetechjohn, that is thetechjaw.com. And of course, I am on pretty much all the social medias at Rob Dunwood. Well, we thank you for being with us. We also thank our brand new boss, Bob. Bob just started backing us on Patreon. Thank you, Bob. Yay. Well, thank Bob, literally, right now. Bob, Bob is the best. Be like Bob. Yes, you've been thanked, Bob, and welcome. Glad to have you. Speaking of Patreon, stick around for our extended show, Good Day Internet. We roll right into it after DTNS wraps up. But reminder, you can get your show live Monday through Friday at 4 p.m. Eastern, 2100 UTC. You can find out more at dailytechnewshow.com slash live. We'll be back tomorrow with a special holiday AMA show, Ask Us Anything, featuring Patrick Verdin and Len Peralta joining us. Talk to you then. This show is part of the Frog Pants Network. Get more at frogpants.com.