 Daily Tech News show is made possible by its listeners, thanks to all of you, including Erwin Sturr, Ken Hayes, and Phillip Shane. Coming up on DTNS, Dr. Nikki is here to explain that big advance for fusion energy you've probably been seeing headlines about, and to calm our fears about general AI replacing humans, even me. This is the Daily Tech News for Tuesday the 13th of December 2022 at Los Angeles I'm Tom Merritt. And from Studio Redwood, I'm Sarah Lane. Brand new from Alabama, I'm Dr. Nikki Ackermans. And on the show's producer, Roger Chen. Yes, we made Dr. Nikki move to Alabama so that we would have more states represented. So that my bird box could be better. And so that we could A-B test your bird box against New York City. No, real quickly, you've got a new job, is that why you're there? Yes, I am a newly assistant professor at the University of Alabama. Well, congratulations. Very cool. Thank you. Thank you. Yeah, congrats. All right, let's get into the quick hits. The US Securities and Exchange Commission has charged FTX co-founder Sam Bankfriend-Freed, we'll call him SBF going forward with fraud for allegedly moving customer funds from FTX to another of his company's Alameda Research. The US Commodity Futures Trading Commission filed a civil lawsuit against SBF for the same reasons, and the US Attorney's Office for the Southern District of New York brought eight criminal counts, including alleged wire fraud, securities fraud, and money laundering. SBF is alleged to have used customer deposits to fund risky investments in Alameda Research, causing both companies to go bankrupt. Authorities in the Bahamas arrested SBF Monday, and he will be extradited to the United States. I just love the way you said, with fraud. I mean, we all kind of knew. Yeah, right. Yeah. What are you going to do? Let's call it fraud. US Senator Marco Rubio of Florida introduced a bill meant to ban TikTok in the United States. The bill would block transactions with any social media influenced by the governments of Russia or China, and that would, of course, probably be applied to bite dance and TikTok. In 2020, the United States Government's Committee on Foreign Investment in the United States ordered China's bite dance to divest itself of TikTok in the US. That never went away. They never made them do it, but talks have been ongoing ever since to address national security concerns. Apple CEO Tim Cook has been visiting Sony manufacturing facilities in Japan posted on Twitter that Apple has partnered with Sony for more than a decade to make camera sensors for the iPhone. Apple had never before confirmed that Sony made that actual part, so they're confirming it now. Apple also announced Tuesday has invested more than $100 billion into its Japanese supply network over the next, over the last five years, highlighting another way that Apple is pivoting away from Chinese manufacturers. Yeah, Tim Cook doing a charm offensive in Japan, you know, talking about how much he loves it, posting on Twitter, all kinds of interesting things going on. The US and the EU used to have an agreement called Privacy Shield that allowed companies to move data back and forth between the two regions while maintaining the privacy protections of the citizens in those regions. But challenges to those agreements have meant that the companies have not had that protection for years now, instead having to each negotiate their own what are called standard contract clauses or SCCs or alternatively just stop transferring customer data between the regions, which increases the cost of your business quite a bit. However, the EU published a draft approval Monday to reestablish a framework for all companies. And this is because in October, the United States began implementing new safeguards on how its intelligence agencies can access such data. So under this new framework, EU citizens will be able to appeal data handling decisions to an arbitration panel. And the United States has agreed to limit intelligence agency data collection. However, lawyer Max Schrems, who is the person who successfully sued to strike down previous Privacy Shield agreements, has indicated that he finds this new agreement insufficient as well. Oh, boy. So it's not over. Probably not. Amazon announced new features for its digital voice assistant gestures on Echo Show devices. Those are the ones with screens and a camera. Let a user interact with the device using just their hands. The only gesture so far, pretty limited, is the ability to hold up your hand palm toward the camera to dismiss a timer. I actually tried to do this earlier and it did not work. Did you turn on the setting though? Uh, no, because gestures are only on by default. So I didn't. They're not on by default. They're, they're not on by default. Yes. You have to turn them on by default. Which I did not do. There's also a new text to speech option that can read out the title of menu options on the screen and consolidated captions turns on captioning for calls or videos or voice responses across all Echo Show devices on a single account. I can't wait to turn this on and stop the timer with my hand. All right, folks, you've been hearing a lot about fusion in the headlines. Dr. Nikki is here to tell us all about it. I'm so excited. I'm the person who gets to talk about this on DTNS today because this is a huge breakthrough. So Tuesday, scientists at the National Ignition Facility at the Lawrence Livermore National Laboratory in San Francisco announced that for the first time ever they achieved a net energy gain during a controlled fusion experiment. So if you've been following fusion science for a while, you know that this is a huge deal. And if not, I'm going to break it down for you a little bit more. So nuclear fusion is the type of energy that powers the sun. Understandably, this is something that we want to harness down here on earth because it would be a great source of green energy if we could make it work. It also doesn't emit carbon dioxide or long live radioactive byproducts like from the nuclear power that we currently use. So researchers have been working on decades to try and achieve stable fusion. And this happens when two hydrogen atoms are mushed together under really high pressure and they fuse and they form a helium atom. And in doing so, they release a bunch of energy. It, however, takes a lot of energy to actually cause this reaction and even more for it to be sustained. So the hope is that in the first round of fusion in a controlled laboratory environment, enough energy is released in the form of heat to trigger a sort of chain reaction with another fusion and so on and so forth, which would be a cascade of fusion reactions that would run until your said fuel runs out. So so far, the amount of energy, sorry, the amount of energy necessary to achieve fusion has outweighed the amount of energy coming out. We've been putting a ton into it, but the results haven't been, you know, as good. Yeah. So you put 10 in, but you only get nine out. That's not. Yeah, or even way less. Right. And, you know, we always hope to get to the point where we would get more out. But some people weren't even that confident to reach this reaction phase, which is called ignition. However, in August last year, the NIF got really close to that threshold. They generated around 1.3 megajoules of electric of not electricity, but energy against the input laser power of 1.9 megajoules. So it's close. But on December 5th of this year, they actually achieved ignition. So yeah, clapping. Excited. Yeah, it's a huge. It's a big deal. It's a really huge deal. And all of this was achieved inside a teeny tiny capsule that's like this big, the size of a peppercorn filled with hydrogen isotopes and a type of fuel made of deuterium and tritium, which is a type of fusion fuel. And this type of fusion is called inertial confinement. So the fusion reaction was kickstarted at 2.05 megajoules of energy coming out of the largest lasers in the country, which is also pretty cool. And they were pointed at this tiny capsule and then they got 3.15 megajoules out, which is 50 more percent of what went in. Nice. This is the accumulation of over 60 years of research across the planet. And obviously it's a huge step towards proving that this type of reactor can actually generate energy. But this was only a proof of concept. So maybe, Tom, you can tell us more about what are the what's going on in the future. Or maybe your dog can tell us more. Ray would like to step in and tell you. But Ray, I'm going to take this one because I agreed to be the downer and explain about the hurdles. It's important to note that the experiment produced more energy than went into the experiment. That's not the same as producing more energy than was used in the experiment overall. These lasers are very inefficient. It took more than 300 megajoules of energy to create the 2.05 megajoules that the laser put out. That's not very efficient. It's still not a headline, though. Yeah, yeah, exactly. It's still scientifically important because what we showed, if it checked me if I'm wrong here, Nicky, is that if we had a perfectly efficient input that you could generate net positive energy. No, that's correct. Yeah, so the science has been demonstrated. It's a matter of now the practicality of making those lasers more efficient. And to do that, we'll have to see if we can repeat the process with more energy efficient lasers. They admitted at Lawrence Livermore, they were just trying to make this happen. They did not use the most efficient lasers that they could use. These are actually kind of old lasers. Or there might be an ability to do this similarly with a different method, like the magnetic devices used at JET or Eiter-Takamak reactor in France. Those are different ways than the what did you call it? The peppercorn. Yeah, different ways than the peppercorn. The constrained nature of doing the capsule, they do a different thing with the magnetics. But the point is, no, this is not going to lead us to diffusion any time soon. But yes, it's huge. And I think it even has a psychological effect to kick off people going, oh, this isn't theoretical anymore. You actually did it. You made more energy come out than you put in. Now we just need to work on how to make that better. Make it like you said, work in a chain reaction so you can have a sustained reaction. And that gets people excited enough to really apply their brains to this sort of thing. Yeah, and to the point that scientists are still warning like, listen, this is really exciting, but don't get too excited. We're still a few decades away from, you know, powering our houses and and businesses with this kind of reactor. And also, fusion is not going to be the immediate response to climate change. Like we need to work on other stuff. We're not going to sit here and wait for fusion to work. So they realize that this is such a big step that people are getting hopeful. But, you know, we still need to focus on other things. But like you said, this is a huge first step. And like we can make lasers more efficient. Like we're better at that than making fusion happen. So right. That was the harder problem. Yeah. So it's super exciting, no matter what. Yeah, I agree. Sarah, what do you think? Well, for anybody, I don't know. I mean, as the layman in this conversation, Nikki, I mean, for anybody who's saying, OK, well, as a consumer, maybe nothing changes tomorrow. But what does change in the next five to 10 years? Well, let's I mean, you know, I'm not a fusion specialist, but let's say, you know, things follow track, the lasers get better, things get more efficient. And in 25 years, they're able to run a fusion reaction in a chain and cleanly and safely produce energy. This could replace most of the things that we're using now. It's better than nuclear fission, which creates radioactive waste. It doesn't release carbon like all the coal burning and gas burning that we do. So we could actually have clean energy and a different clean energy than what we have today with windmills and things like that. So but that's a lot of ifs to get there. And I don't know if that's going to impact like prices. I don't know. The lasers are really, really energy consuming to start up. So what does that do price wise? I don't know. But yeah, that's the other part of this. It's the practicality of of making the laser, the input more efficient, whether you're using lasers or whether you make a tukamak work or something else, you're going to have to bring that down. And then there's cost. That's a whole separate issue. You might be able to do a really efficient input that makes fusion work and generate more electricity than is even used in the entire system, which would be great. And it costs a billion dollars a month to run. And that that's not going to be very practical economically. So there's that side of it, too, right? Yeah. And like you said, with the magnet one, for example, that's another possibility. I know that those require even more energy than the laser one. So it's kind of a race to see who's going to go first. But I'm I'm here and we're going to keep watching it and covering it on the show. I'm sure. Yeah. I mean, we're a little bit in the world of, you know, Goddard was able to make a rocket go up out of the atmosphere. And there was it was a big long step before you could actually put people on a rocket. But eventually we got there and then we sped through putting people onto the moon and then having reusable rockets and shuttles and all kinds of things. So it's, you know, we're we're early on in the process, but we're in the process where we went from we think this could work someday to, oh, we made it work. Now we just have to make it practical. It's from maybe not in our lifetime to maybe in our lifetime. So yeah. Yeah. Yeah. And then what are we going to do with all that helium? Yeah, we don't need it. Well, what we do need is for all y'all to give us feedback about the show. One way to do that is in our subreddit, submit stories and vote on them at dailytechnewshow.reddit.com. News about AI. People use it as a catch all for everything these days. But you see it in the headlines. The news comes fast and furious these days. TechCrunch just now reporting SensorTower data showing that that Lenza AI app, the one that is responsible for all the weird looking avatars people use on social media, so many of them. I have a few myself. It not only is continuing to top app store charts, but it's followed by several imitators, including three other AI art generators in the top 10. Do not pay demonstrated a chat bot that can interact with customer support chat at companies often handled by actual humans to negotiate better rates on like your cable bill or your phone bill or even cancel service. Those are stories that just came out today. You could be forgiven for thinking that general AI, the kind that people worry about taking over the world is just about here. So, Nikki, can you explain, first of all, what we mean when we say general AI and why I shouldn't worry that Skynet is going to arrive soon and kill us all? Yeah, Tom, I can I can explain. And I've got to say the reason this this topic came up is like it's a debate in my family all the time. They're like, no, I'm sure the robots are coming. Like every time something new comes out, they're like, they're going to get us. And I'm like, well, first of all, general AI is generalized. AI is is like really real intelligence. And and the fact that a computer could have its own thought that's not caused by humans inputting parameters or a data set. And we are not there yet. So the main problem with AI right now is that it's unreliable in unusual circumstances, because we are training them to be good at repeating things that have already been done. It can't really come up with new things very well. So it has a prompt. It follows the prompt based on the mathematical likeness of whether X or Y answer is more correct according to its data set. An example would be with Dolly, the artist in quotes, AI. And Dolly isn't able to tell the difference between the phrases an astronaut riding a horse and a horse riding an astronaut because it makes the mathematical assumption that one of them is more likely to be used than the other. And that's something that a human mind obviously could very easily tease apart. Yeah, we generally don't see horses riding astronauts. We just kind of got know that, right? Yeah. Yeah. Don't need to be trained. You could draw that if you wanted to. Yeah. So and then to follow up things like chat GBT that we've been talking about on the show a lot, they seem impressive, but it's kind of almost only a surface level. And that's because as with most of the machine learning processes, it's based on a data set. And this data set could be super large, but it's not infallible. And the calculations used to produce content can only run based on the parameters it's given by the developers. So original ideas couldn't be original if they followed a pattern. And that's why these things, these AIs can't come up with original ideas. And all the prompts created by chat GPT for like film scripts sound like an infinite iteration of Hallmark movies. They're kind of basic. Yeah. Chat GPT, you can almost always see where it got its ideas, even if it's not copying them, although sometimes I've noticed that it just copies them, but you could you're like, oh, it's taking this piece, this piece, this piece. If you were to say to chat, chat GPT, come up with an idea for a film script that has never been done before. It's going to fail. It's going to come up with something, but it's not going to come up with a film entitled never been done before. Right. Or something like that. Movie. Right. And it'll be a film. It will be a different text than has ever been done before. Possibly. But barely. But it won't be a new idea. That's the difference, right? If I'm getting this right. Yeah. And the problem, and I think what a lot of people forget is that the data and the parameters that are restricting this data are human generated. You have to tell the machine what to do. So the AI is automatically going to be biased towards the same human biases that the human programmers are. And we've covered biases extensively on a previous episode. But, you know, you can imagine that if you have a huge data set of human faces and you want it to recreate a certain face, it's going to under-represent whoever wasn't able to submit their face into the data set because they don't have an internet connection or something. Just things like this, just kind of snowball with AI. Yeah. So AI, machine learning, all these things are advancing at a huge breakneck speed and at a much faster rate than regulations are able to keep in check. We already know that our laws are slow, but compared to something like how AI is advancing, they're like snails. And so it's really important that we keep them in check and we keep kind of on our minds the benefits versus the drawbacks that we can predict for AI user cases. We don't do we need an AI that tries to clone dinosaurs? Probably not. Like we can focus them on things that are useful. And do you, Nikki, you know, as somebody in the science community feel like you're getting pushed back on things like this? Um, no, I mean, I, I feel like my peers kind of know that like right now it's not a big deal. There has been an AI that writes research papers and it's very funny to read. It's, it's like, I think one of the questions they submit to it was do vaccines cause autism? And the AI said yes and no. And so it's like, because it's probably found data on the internet that said yes, right? It said both. Yeah. Yeah. So I'm not threatened yet. That's what I'm saying. And the cool thing about these AIs, and I say AI generally, but yeah, it's machine learning and algorithms. There's a lot of inflammatory headlines that say that, you know, it's the end of the world, et cetera. But there is a silver lining. These products are in specific instances really, really useful for stuff that like we're too lazy to do. We don't want to do really complicated max math problems over and over and over. And my favorite example is actually one from Scott Johnson and another colleague on the show who says something like Dolly can help create character inspiration for artists instead of trying to find the perfect Google image. Just type it into the prompt and then you can draw from there. Yeah. And no one thought that was going to be what it was going to be used for. I think that's what we're experiencing with a lot of these different machine generated tools is that at first they're so impressive and maybe a little beyond what we expected that we get freaked out or easily impressed or both. Then, like we have with all technology, it's pretty soon we get used to it and start picking it apart. Like, oh, it's bad at this, doesn't do this well. Look at this horrible, didn't even put a head on my my Lenza AI avatar. You know, that's what we do. And so I love that humans do that. I think as we become more familiar with these tools and they come out in the public more often, we'll become less freaked out about this. What I'm curious is what are the signs that something is a generalized AI? Like and granted, as you've I think well demonstrated, that's probably farther down the road than fusion reactors. Like, what would it look like, though? What would it do? I think if it gets to the point where the majority of the population can be tricked by it or even people who just aren't aware that it's an AI can't tell if it can pass its own by a more specialized Turing test, like not the Turing test, but like a more specialized Turing test. If it passes it, then we need to start thinking about that. But we're not there yet. For now, it's it's not like all the AIs that we have would need to be all grouped together in one like mega AI and then be better and then we could get closer. But they all do very specific tasks. Right, because the chat GPT won't play Stratigo very well. But the Stratigo AI is not going to be able to create art and the lens AI can't create a film strip. But even if you combine them all together, you're still going to find gaps of like, well, this can't do that. There's there's the other thing I want to do. And in general, I may not do everything well, because humans are general intelligence. We don't each of us doesn't do everything well, but we would be able to, you know, conceive of it and come up with a, you know, an approximation of an answer for anything. Yeah. And we can also go super philosophical and be like, well, what is intelligence? Like maybe this is AI's own form of intelligence. Like chickens are intelligent, but they can't they don't have thumbs. So they can't use tools. So like, what are we, where do we want to go with this? Well, and that would be really interesting is if an algorithm at some point demonstrates something that looks like intelligence that doesn't look like anything humans, dogs, dolphins or anything. I'd be excited for that. Like maybe that could be super helpful. Yeah, yeah. Unanticipated consequences could be positive. They don't look at the silver lining. Yeah, yeah. Well, speaking of maybe something that nobody asked for, Jeff Bezos's space tourism company, Blue Origin, will release a kid-friendly animated space adventure series called Blue Origins Space Rangers produced by companies, Genius Brands and SMAC Productions. But our release is DVD. They've just said they're new in it. The series will feature the voices of Bezos himself. And good morning, America co-host Michael Strahan, who was a passenger in December of 2021 on Blue Origin NS 19 on a 10 minute space flight. Bezos took his first ride to space in July of 2021. According to the press release, Blue Origin Space Rangers, the show will also include voice talent and other guest stars like junior astronauts, celebrities and adult leaders. So what I'm unclear on from what they've released about this is if Bezos is playing himself or not. Also, what's an adult leader? Oh, I mean, I don't know that to me, that just means like. Camp counselor. Yeah, they're going to they're going to pick politicians and heads of the G.O. Sure, like that. I sure that that's just that's a catch all term for important people who aren't celebrities or like a big weird ego trip. I'm sorry. I would like to say I'm an adult leader. OK, well, sure. Yeah, I mean, that was just in the Yahoo story. I don't know if that's the the actual text that Amazon is using to make it in the press release. Yeah, yeah, I can't hire like actual astronauts to do this. Why does that? Well, they are. They're going to have they're going to have actual astronauts in here. And in fact, Michael Strahan qualifies as an astronaut because he's done it. He's done it. I just like the fact that cartoons used to promote selling toys. And now they promote aerospace industrial projects. Do your parents want to go to space for three hundred thousand dollars? They can't instead of bugging mommy and daddy to buy that toy in the toy aisle, bug mommy and daddy to lobby their Congressperson on behalf of Blue Origin. Get y'all to Mars. All right, let's check out the bail bag. All right. So Terry wrote in about our discussion about terms of service with IA's act are yesterday, Terry said, it occurred to me that a simple solution could be put forth by either a big three group or the Fed having standardized abbreviated terms of services that could then be referenced. IE the general EULA could be a general international EULA. The use terms could be generalized and referenced as such. The liability waiver could also be generalized. Then to invoke each package, there could be a checkbox or a bulletless saying, quote, to use our service, you agree to the federal regulation EULA terms of use or federal regulation liability waiver. The following terms are specific to our product. That is what Terry had suggested. He also says below listed would be this specific terms of service for anything not covered under the regulated terms of service scripts. We will add that to our show notes. Yeah, we'll have the email in the show notes. He's saying below the checkboxes thing, they then every company could have anything else that wasn't covered by the by the the object oriented thing. I think this is an interesting idea, sort of, you know, object oriented. If your objection to this idea is like, well, I don't want the government involved. And OK, let's make it an industry standards organization, right? Similar to USB or something where we have blocks of of EULA content that everybody knows what it means. Everybody agrees what it means. The industry agrees that it protects them. I think the the fact that you would have specific terms of service for every company does lend a little backdoor to them putting in whatever they want, but at least it's less to cover and audit. I don't know, folks, what do you think of this? Let us know feedback at Daily Tech News Show dot com. In deed. Also, thanks to you, Dr. Nicky Ackermans for being with us today. Let folks know where they can keep up with your work. Yeah, they can go to my website is Nicole Ackermans dot com or I'm still on Twitter for now at Ackermans Nicole until it dies. Some I'm going down with the ship. Well, it's not dead yet. But yes, yes, many places to find various folks. So that's where you can find Dr. Nicky very happy to have you with us, though. Also very happy to have a brand new boss and that boss's name is Luke. Luke just started backing us on Patreon. Thank you, Luke. We're so glad to have you. Indeed. Thank you, Luke. Luke is now immediately signed on for the new Patreon merch. If Luke sticks around for three months, unless they're at the production assistant level, all the other levels will get a piece of merch with Len Peralta's exclusive nine year anniversary art on it. So at the various levels, there's stickers, there's t-shirts, there's hoodies, there's tote bags, there's coffee mugs. If you're not a patron already, now's a good time to get started. Patreon dot com slash DTNS speaking up, Patron, stick around for our extended show, Good Day Internet, where we talk about all the things. You can also catch this show. DTNS is live Monday through Friday at 4 p.m. Eastern 2100 UTC. Find out more at dailytechnewshow.com slash live. We are back tomorrow with Scott Johnson and Will Smith, and we're talking PC upgrades. See you then. This show is part of the Frog Pants Network. Get more at frogpants.com. Diamond Club hopes you have enjoyed this program.