Tech Brew Ride Home - Fri. 12/22 – The Hyperloop Is Dead

Episode Date: December 22, 2023

The Hyperloop dream is dead. The attempt to bridge the iMessage divide is dead. China’s gaming industry isn’t dead, but it’s pretty wounded. The crazy story of that teenage hacker who leaked the GTA VI stuff. The weekend longreads suggestions, and at the very end, a unique look back at the evolution of the AI moment. Sponsors: Nutrafol.com/men code ridehome Links: The hyperloop is dead for real this time (The Verge) Beeper is giving up on its iMessage dream (The Verge) Apple’s Newest Headache: An App That Upended Its Control Over Messaging (NYTimes) Tencent Leads $80 Billion Rout as China Rekindles Crackdown Fear (Bloomberg) Lapsus$: GTA 6 hacker handed indefinite hospital order (BBC) Weekend Longreads Suggestions: The Race to Fill Crypto’s FTX-Shaped Hole (Wired) As the AI era begins, Reddit is leaning into its humanity (Fast Company) Tolkien Estate Wins Court Order to Destroy Fan’s ‘Lord of the Rings’ Sequel (NYTimes) Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the Tech meme right home for Friday, December 22nd, 2023. I'm Brian McCullough today. The Hyperloop dream is dead. The attempt to bridge the I-Message divide is dead. China's gaming industry isn't dead, but it's gotten a pretty significant flesh wound again. The crazy story of that teenage hacker who leaked the Grand Theft Auto Six stuff and the weekend long read suggestions and at the very end, a unique look back at the evolution of the AI moment. Here's what she missed today in the world of tech. The Hyperloop is dead. Yes, sometimes even Elon Musk has a failed business venture. Quoting the verge, Hyperloop 1 is selling off its assets, closing down its offices, and laying off employees.

It will formally close at the end of the year, at which point all of its intellectual property will shift to its majority stakeholder, major Dubai port operator DP World. Whoever buys the test track in the Nevada desert will have one hell of a slip-and-slide if they want it. Since its founding in 2014, the company raised around $450 million in venture capital funds and other investments. While there is still a small smattering of startups trying to build hyperloups, the demise of one of the biggest hyperloop companies signals the end of the dream that originated with Elon Musk's so-called alpha paper in 2013, end quote. Elon's vision was for a fifth mode of transportation involving sleek, aerodynamic aluminum capsules capable of transporting passengers or cargo at speeds of. up to 760 miles per hour through almost vacuum-like tubes. These tubes, potentially elevated on pylons or buried underground, were supposed to connect or exist within cities. People were especially keen on

Musk's proposal of reducing the Los Angeles to San Francisco journey to a mere 30 minutes, sparking widespread interest among engineers and investors globally. Originally named Hyperloop Technologies, the company later became Hyperloop 1 in 2016 and eventually Virgin Hyperloop 1, following its acquisition by Richard Branson. But financial challenges were a constant hurdle. Richard Branson's intervention brought a crucial $50 million investment from two existing stakeholders, aiding in meeting immediate financial obligations. In 2019, the company raised a significant $172 million with D.P World, a prior investor contributing $90 million and holding two board seats. Virgin Hyperloop 1 did construct a Nevada test track to assess the technology safety and viability.

In 2020, it achieved a milestone by conducting its first huge. human passenger test, albeit reaching only 100 miles per hour significantly below the ambitious initial target. Despite technical plausibility, the Hyperloop concept faced skepticism, labeled as a utopian vision with questionable financial feasibility and criticized as perpetually just around the corner. In 2017, version Hyperloop executives optimistically predicted operational hyperloups worldwide by 2020, a timeline later extended to 2021. The COVID-19 pandemic further impacted the project leading to the departure of top executives and founders from Hyperloop 1, which also dropped Virgin from its name, shifting its focus from passenger to cargo transport.

As of now, a fully operational hyperloop system does not exist anywhere in the world. Musk's own test tunnel in California has been dismantled. His boring company continues to develop underground passageways in Las Vegas, but these are designed for Tesla's, not Hyperloops. Remember Beeper and their grand dream I told you about the dream of the dream of of breaking down the blue-bubble-green-bubble barrier. Well, like the hyperloop, it looks like the dream is dead, quoting the verge. Although it just announced yet another fix to let Android users send iMessages from its messaging apps,

Beeper announced that it's going to stop finding workarounds and instead focus on its, quote, long-term goal of building the best chat app on Earth. Each time that Beeper Mini goes down or is made to be unreliable due to interference by Apple, Beeper's credibility takes a hit. It's unsustainable, Beeper writes, as much as we want to fight for what we believe is a fantastic product that really should exist, the truth is that we can't win a cat and mouse game with the largest company on Earth, end quote. What started as a simple app download in Beeper Mini has become an increasingly complex process for Beeper users,

and its latest fix seems like its most desperate attempt yet. Beeper wants users to own or rent a jailbroken iPhone, along with having a Mac or Linux computer. Users can start using the app once they follow Beeper's steps to receive their iMessage registration code, the only problem is that they'll need to leave the iPhone plugged into power and connected to Wi-Fi at all times. If users don't own an old iPhone or don't want to purchase one, Beeper suggests renting one for, quote, a few dollars per month. The company says that it will offer its own rental service to do so next year, quote, if there is enough interest. But if Apple responds with a way to block Beeper's latest method, Beeper says it does, quote, not have any current plans to respond if this solution is knocked offline, end quote. Yeah, maybe I'm just spitballing, but maybe, you know, going toe to toe with the biggest

richest company in the world is a losing proposition. But sources are suggesting that Apple blocking Beeper Mini has caught the attention of the DOJ, which is in the middle of a four-year-old probe into Apple's anti-competitive behavior, quoting the New York Times. Since it was introduced on December 5th, Beeper Mini has quickly become a headache and potential antitrust problem for Apple. It has poked a hole in Apple's messaging system, while critics say it has demonstrated how Apple bullies potential competitors. Apple was caught by surprise when Beeper Mini gave Android devices access to its modern iPhone-only service. Apple's reaction set off a game of whack-a-mole with Beeper Mini finding alternative ways to operate and Apple finding new ways to block the app in response.

The duel has raised questions in Washington about whether Apple has used its market dominance over iMessage to block competition and force consumers to spend more on iPhones than lower-priced alternatives. The Justice Department has taken interest in the case. Beeper Mini met with the department's anti-trust lawyers on December 12th, two people familiar with the meeting said. Eric Mijikovsky, the co-founder of the app's parent company Beeper, declined to comment on the meeting, but the department is in the middle of a four-year-old investigation into Apple's anti-competitive behavior. The Federal Trade Commission said in a blog post on Thursday that it would scrutinize, quote, dominant players that, quote, use privacy and security as a

justification to disallow interoperability between services. The post did not name any companies, end quote. Maybe it's a losing game to try to battle Cupertino, but it's also possibly a losing game to try to fight Beijing. We've talked about bite dance growing bigger than meta. We've talked about fast fashion e-commerce outfits like Sheehan, taking a bite out of Etsy's market. But if you're a Chinese startup and you grow too big, you always run the risk of the Chinese government pulling the rug out from under you. The Chinese government has published draft rules for online gaming, including setting spending limits and banning daily login rewards. This has caused the market cap of Tencent to plunge more than 10 percent and net ease more than 20 percent.

Quoting Bloomberg. The sweeping restrictions which caught industry players and investors off guard on the final trading day before Christmas reminded many of the brutal technology sector crackdown of 2021. That year, various agencies abruptly imposed curbs on sectors from e-commerce to entertainment, reining in Jack Ma-backed Ant Group and Alibaba Group holding while decimating the online education industry by declaring profits illegal. As with two years ago, Friday's regulations emerged with little warning and were at once so vague and all-encompassing that investors couldn't decipher the intent or potential fallout. Outraged and confused posts dominated a WiiChat group of hundreds of developers and designers, many complaining in particular about the unspecified cap on player spending.

Chinese games are known for shrewd designs and promos that encourage players to spend on decorating and burnishing their avatars, the main source of income for Tencent, and a its rivals. The Communist Party since 2020 has waged a campaign against a private sector it regarded as amassing too much power and expanding recklessly, threatening its control of the world's number two economy. Xi Jinping's administration has long sought to combat gaming addiction, blaming online entertainment for the rise of myopia among youths. Chinese users spend more time online on average than in many other markets, fueling the rise of services like Doyen and WeChat. Critics have also linked that to various ills from unemployment to low birth rates.

end quote. Do you remember that notorious Lapsis group from a couple years ago? They were the ones who originally leaked the first images of Grand Theft Auto 6. They also hacked Nvidia and Uber and others, really high-profile intrusions into some seriously secure organizations like BT in the UK. Well, in the UK, a judge has sentenced Arian Cortage, one of the Lapsis hackers, to a secure hospital until a mental health tribunal lets him leave. Kirtage, who is apparently autistic, will remain in the hospital for life unless doctors consider him no longer a danger to society. What kind of a danger to society?

Well, get this detail from the BBC, quote, a mental health assessment used as part of the sentencing hearing said he continued to express the intent to return to cybercrime as soon as possible. He is highly motivated, end quote. The jury was told that while he was on, bail for hacking Nvidia and BT and in police protection at a Travelodge hotel, he continued hacking and carried out his most infamous hack. Despite having his laptop confiscated, Kirtage managed to breach Rockstar, the company behind GTA, using an Amazon firestick, his hotel TV, and a mobile

phone. Kirtage stole 90 clips of the unreleased and hugely anticipated Grand Theft Auto 6. He broke into the company's internal Slack messaging system to declare, if Rockstar does not contact me on Telegram within 24 hours, I will start releasing the source code. He then posted the clips and source code on a forum under the username T-Pot Uber Hacker. He was re-arrested and detained until his trial, end quote. As Dar Obasanjo said on threads, quote, I'm barely able to connect my Netflix to a hotel TV, and this dude used his to hack Rockstar games, log into their company Slack,

and brag about hacking them, and then leaked clips of GTA-6. Mind-blown emoji. In a six-week trial at Southwark Crown Court, a 17-year-old member of the hacker group Lapsis was also found guilty of cybercrimes alongside Cortage. This group attempted to extort $4 million in ransom from various targeted corporations. Additionally, they targeted individual cryptocurrency wallets. The 17-year-old received an 18-month youth rehabilitation order, which includes intensive supervision and and a prohibition on VPN use. He was also sentenced for engaging in a disturbing pattern of stalking and harassment against two young women. Kurtage and this minor are the first convicted members of Lapsis, a primarily teenage hacker collective, apparently in Britain and Brazil. The total financial

gains of Lapsis remain unclear, as no companies have admitted to paying the ransoms, and the hackers haven't released passwords for the seized cryptocurrency wallets. Time for the week on Long Read suggestions. first up. When FTCX blew up, it actually blew a hole in the crypto market. And with Binance, also having its issues recently, Wired takes a look at the race to fill that hole in that market with some new companies. Quote, there are a lot of exchanges even with FTCS out of the picture, says Venture Investor William Quigley, who also co-founded the Tether Stablecoin. But there is an opening, he says, for an exchange that can demonstrate it stores customer assets responsibly, protects against market manipulation, and follows rigorous compliance procedures.

That's an area ripe for improvement, Quigley says. New players are pitching more technically elaborate methods of proving that customer funds have not been F-TX'd. Backpack is developing a new proof of reserves, updated automatically on a daily basis, explains Ferranti, whereby the availability of funds for withdrawal is demonstrated cryptographically as opposed to through an opaque internal audit. To prevent funds from being quietly shifted about, the exchange will operate under a system whereby each crypto token transfer must be authorized by multiple parties. The aim is to ensure there is no single point of failure, says Ferranti, and the exchange has multiple levels of of defense. Other competitors like OpNX, an exchange launched in April by Kyle Davies and Suju,

the co-founders of Bankrupt Crypto Hedge Fund, Three Arrow's Capital, are trying to scoop up former FDX customers with a different approach. OpinX provides regular crypto trading, but also lets customers trade their bankruptcy claims. Instead of waiting out a lengthy bankruptcy process, someone with money locked up on FDX could choose to sell off their claim for a certain number of cents on the dollar, swapping maximum recovery potential for immediate access to funds, end quote. Then in Fast Company, Harry McCracken takes a look back at the whole Reddit brew ha-ha, as Reddit pivots to the new year, pivots hopefully to an IPO, and also pivots to AI by embracing its humanity, quote. Ugly, though, the moderator rebellion was it reflected Reddit's greatest strength, the deep engagement

of its members. Over 50,000 moderators preside over subredits visited by 70 million daily active users. The 16 billion posts and comments they've created, mostly without profit as a motive, make up one of the Internet's most essential troves of unvarnished real-world knowledge. The API controversy also danced around the edges of emerging questions that may turn out to be vital to Reddit's fate as a public company. For years, one of the best ways to get savvy answers to nettle some everyday problems has been to perform a Google search and click on any Reddit links in the results. While I was working on this story that helped me troubleshoot my flaky garage door opener

after the manufacturer's site proved useless. But what happens if conventional Googling gives way to people getting their answers directly from chatbots, or if lowest common denominator AI-generated content overwhelms the Internet in general? Can Reddit ensure that its foundational humanity remains an asset rather than an albatross? The answers are yet to come, but the tension is already in the air. Underneath it all, Reddit does have the bones of one of the Internet's oldest, sturdiest institutions, the message board. Hundreds of thousands of them in this case, where Redditors can share links and opinions,

answer each other's questions, or just indulge in memes and other classic forms of online play. The fundamental appeal of that proposition has made Reddit one of the U.S.'s top 10 most trafficked websites and among the top 20 internationally, according to similar web. According to Huffman, Reddit's secret of future success, mostly requires nailing the basics in ever more ambitious fashion, including continued efforts to ensure the platform appeals to the masses, not just its most committed fans, end quote. And finally today, I knew nothing about this. Did you know a fan made an unauthorized Lord of the Rings sequel? How'd that go? quoting the New York Times. The Tolkien Estate learned that the book had been published in March 23 when it was being sold through Amazon and Barnes & Noble. The estate said in court documents that it sent Polycron a

cease-and-assist letter and tried to reach him several times by phone to no avail. In April, Polycron sued the Tolkien Estate and Amazon. He claimed that Lord of the Rings, Rings of Power in Amazon Prime Video Prequel series that was released last year and is one of the few adaptations authorized by the Tolkien Estate infringed on the copyright of his book. He asked for $250 million in compensation. That case was dismissed in August with Judge Wilson finding that Polycrown's book was infringing on the estate's copyright. In a summary judgment, Judge Wilson found, quote, direct evidence of copying and borrowed Polycrown from further distributing the book or any others in a planned series. He also ordered Polycrown to destroy all electronic and physical copies of the published work, The Fellowship of the King,

by Sunday. As of Wednesday, Amazon and Barnes & Noble were no longer listing the book for sale online. end quote. This has weirdly been a day of stories about David's battling Goliath and not quite working out for the Davids. Okay, so obviously no show on Monday because it's Christmas, but I think I'm going to have something for you every day that I'm gone. One thing for sure is that tomorrow I'm going to post an appearance I made on the Newsworthy podcast recently. Since I didn't have time to schedule a year-end tech trends roundup for 2023 for this show, I asked and got permission to repost the TechTrans episode I did for them. So that's coming tomorrow. Then I'm going to try to find some internet history episodes to give you probably Sunday and definitely Monday, but I haven't

picked those out yet. And then, right now, you know that I don't like to do more than two ads on a show if I can avoid it, but you've also heard Miro a lot recently. They bought a bunch of ads at the tail end of this year. So many, in fact, that I missed airing one of them. So I'm going to play a make-good ad for them right now. And to make up for it, I'm going to give you something I've been sitting on for a while. On July 20th, I did a segment on this show, July 20th, 20th, 20-20, talking about Open AI and the release of GPT-3 for the very first time. It's about five minutes long, and it's wild to hear three and a half years on how much things have evolved from that nascent state. So, first, the make-good ad, enjoy the look-back.

in AI history and I will talk to you on Tuesday. Finally today, let me hip you to something that got a lot of attention over the weekend. It's not really a news story per se, but let's just say a lot of folk in Silicon Valley have woken up and decided that GPT3 might herald the beginnings, the stirrings perhaps of a next big thing in tech. Let me give you some background. Open AI is an AI research Foundation, started by Elon Musk, Sam Altman, as well as a who's who of big names and things like machine learning. GPT3 is OpenAI's text-generating AI that debuted about a month ago and is a language model trained with 175 billion parameters. It's a successor to the previous GPT2, which had only 1.5 billion parameters. In short, remember how we've discussed on the show before, how AI can't

really read, can't really speak, and that's sort of holding it back? Yeah, well, this is from an essay by Delian Asperuhoff that's been getting passed around all weekend. Delian has been playing with GPT3, and let's just say he was impressed. Quote, GPT3 is essentially a context-based generative AI. What this means is that when the AI is given some sort of context, it then tries to fill in the rest. If you give it the first half of a script, for example, it will continue the script. Give it the first half of an essay. It will generate the rest of the essay. I fed it half of an essay on how to run effective board meetings. And it effing wrote up a three-step process on how to recruit board members that I should honestly now put into my damn original essay. Both times the AI was

able to generate cogent, although not always correct, additional paragraphs, and in both examples, was able to follow the prior formatting, i.e. section headers and steps. What's incredible about the tool is that you can feed it almost any context, a script about a gay couple in Italy, an interview between two tech luminaries, or even a political column about an election, and it is able to put together decently coherent arguments. Now, before you get too excited, this isn't some sort of general AI and the machine doesn't really have a way of understanding if what it is outputting is true or not. The simplest way to explain how it works is that it analyzes a massive sample of text on the internet and learns to predict what words come next in a sentence

or given prior context. Based on the context you give it, it responds to you with what it believes is the statistically most likely thing based on learning from all this text data. This is a strategy that OpenAI and other researchers have been pursuing for quite some time by starting off with a simple problem like trying to predict the next word in a sentence. We have now steadily built up to where they are today, where a model like GPT3 can complete several paragraphs or more. Though an incredible result, even GPT3 at some point may lose direction and wander aimlessly. Despite its massive size, over 175 billion parameters, as mentioned, it still may struggle with keeping a long-term destination in mind or holding logical,

consistent context over many paragraphs. This means, in my opinion, although there's debate on this, that while this tool is very impressive and GPT4 will likely show further improvements, there are probably diminishing marginal returns to this approach. They can keep getting better at running really complicated statistics on all the text people I've ever written, but the AI is still not capable of reasoning, end quote. Okay, so maybe the AI is not reading and writing just yet, but it has gotten really good at tricking us into believing it's reading and writing. Actually, there's another link towards the bottom of the show notes where Kevin Lacker gave GPT3 a Turing test. It didn't pass exactly, though, if you had run this test maybe 10 years ago,

it would have fooled a lot of people. And the overall results were interesting. I'm not going to read you the whole thing, but I think I can sum up this story by saying this type of AI has made a leap that is exciting a lot of people, because at the very least, this sort of thing should lead to way, way, like generationally, way better voice assistance. And aside from that, there are some interesting possibilities for things like coding and design. Also in the show notes, there are two tweets with live video from Sharif Shamim, where first, he wrote a sentence describing what Google's homepage should look like, and GPT3 generated the code that generated an actual nearly identical identical Google homepage. Then he built a functioning React app by again just describing what he

wanted GPT3 to do and it generated it. All of the code. There's been a bunch of debate and pushback on all of this. Sam Altman himself tweeted over the weekend. The GPT3 hype is way too much. It's impressive, thanks for the nice compliments, but it still has serious weaknesses and sometimes makes very silly mistakes. AI is going to change the world, but GPT3 is just a very early glimpse. We have a lot to still figure out, end quote. But as Paul Graham tweeted also this weekend, hackers are fascinated by GPT3, and to everyone else it seems like a toy. Pattern seem familiar to anyone, end quote.