Tech Brew Ride Home - Wed. 08/17 – “Anti Party” Technology

Episode Date: August 17, 2022

Cracking down on house parties as a service. Apple makes moves to diversify the supply chain. Another worrying datapoint from the semiconductor industry. Have the banks been using WhatsApp to avoid scrutiny from regulators? And can large language models make our dreams of robot butlers a reality? Sponsors: KeeperSecurity.com/techmeme Manscaped.com code RIDE for 20% off Links: Exclusive: Airbnb rolls out new anti-party tech to prevent unapproved gatherings (Fast Company) MacBook design changes enable Vietnamese production for the first time; Apple Watch too (9to5Mac) Chipmakers Are Flashing More Warnings on the Global Economy (Bloomberg) Wall Street's Record Fines Over WhatsApp Use Were Years in the Making (Bloomberg) Amazon is raising seller fees for the holidays to manage through surging inflation (CNBC) Mid-year Crypto Crime Update: Illicit Activity Falls With Rest of Market, With Some Notable Exceptions (Chainalysis) Google invested a whopping $1.5B into blockchain companies since September (CoinTelegraph) Google’s New Robot Learned to Take Orders by Scraping the Web (Wired) 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 Wednesday, August 17th, 2020. I'm Brian McCullough today, cracking down on house parties as a service. Apple makes moves to diversify the supply chain, another worrying data point from the semiconductor industry.

Have the banks been using WhatsApp to avoid scrutiny from regulators? And can large language models make our dreams of robot butlers a reality? Here's what you miss today in the world of tech. Airbnb is introducing what it calls anti-party technology in the U.S. and Canada to automate. automatically block reservations at a high risk of leading to unauthorized house parties. Quoting Fast Company. The vacation rental company tells Fast Company the new tech is aimed at identifying potentially high-risk reservations. Airbnb will be looking at factors like a user's history of positive reviews and history with the service length of the requested booking distance from their

home to the booking and whether it's booked for a weekend or weekday. Short-term rental platforms like Airbnb and VRBO have long struggled with unauthorized parties taking place at bookings. The new tech is meant to flag and prevent suspect bookings from being sent to the host for approval, so homeowners can approve bookings without fear of property damage or noise complaints. People who have been blocked from booking an entire listing, because of the new system, will still be able to book a single room and a whole residence, since the host is more likely to be on the premises or a hotel room. Airbnb has been testing the tech in Australia since October 2021 and says it's seen a 35% drop in unauthorized parties in areas where the pilot was in effect. The safety tool has rolled out nationwide

in Australia and will now be tested in the U.S. and Canada. The latest update builds off the company's under 25 system, which blocks users under the age of 25 from booking entire properties close to where they live until they have at least three positive reviews. The company said in a statement that the new system is meant to prevent more party-seeking users from booking. quote, while having less of a blunt impact on guests who are not trying to throw a party, end quote. In 2019, Airbnb announced a ban on party houses, homes that are essentially listed for the sole purpose of hosting events, and announced a slew of safety features after five people were killed in a shooting that took place at an Airbnb. In August 2020, as the pandemic spread across the globe, the company enacted a full ban on parties, end quote.

A couple of interesting takes on this. Johanna Boyan tweeted, quote, Sounds like predictive policing for Airbnb guests, very vulnerable to bias, end quote. Yeah, find yourself frustrated by the songs or the posts that the algorithms don't allow you to see. Now imagine minority report style algorithms circumscribing behavior in the real world. And Darry Obasandro tweeted, quote, Airbnb is deploying anti-party technology and just like that we've become our parents, end quote. Indeed, insert Beastie Boys joke here, fight for your right, and all that. A couple of stories now showing that the current tech narratives all remain unbroken this summer,

as I asked Noah Smith last weekend, if there's a non-zero chance that geopolitics could shut down your entire business, doesn't it behoove you to do something about it ASAP? Sources are saying that Apple is in talks with its suppliers to mass-produce Apple Watch and MacBooks in Vietnam for the first time. Quoting 9 to 5 Mac. Niki Asia reports that Apple plans to arrange for both MacBook and Apple Watch Assembly in Vietnam, with HomePod likely to follow. Apple is in talks to make Apple Watches and MacBooks in Vietnam for the first time, marking a further win for the Southeast Asian country

as the U.S. Tech Giant looks to diversify production away from China. Apple suppliers' Lux Share Precision Industry and Foxcon have started test production of the Apple Watch in northern Vietnam with the aim of producing the device outside of China for the very first time, three people with direct knowledge of the are told Niki Asia. On the MacBook front, Apple has asked suppliers to set up a test production line in Vietnam. Two sources said, one of the people with direct knowledge of the situation said, quote, the components for MacBooks have become more modularized than in the past, which makes it

easier to produce the laptops outside of China, end quote. Niki says that Apple has increased its number of Vietnamese suppliers from 14 in 2018 to at least 22 today. Difficulties do still remain. Sources suggest that MacBook assembly costs are higher in Vietnam, then China, and Apple Watch Assembly is technically challenging due to the need to squeeze so many components into such a small and packed space. But the country's tech sector is working on improving its capabilities motivated not just by Apple, but also by other tech giants seeking to move production there for similar reasons. Amazon, Dell, and Google have all set up production in Vietnam, says the report, end quote. And another data point that looks bad for the chip industry,

according to trend force, DRAM bit growth, a key barometer for for chip market demand that measures memory produced will rise 8.3% in 2023 the weakest on record, quoting Bloomberg. Memory chips are among the most vulnerable segments in the $500 billion semiconductor market to global economic performance, and Samsung and S.K. Heinix sales of dynamic random access memory or DRAM, a chip that holds bits of data, are central to Korean trade. Next year, demand for DRAM is likely to rise 8.3% the weakest bit growth on record, says tech researcher trend force, which sees supply climbing 14.1%. Bit growth refers to the amount of memory produced and serves as a key barometer for global market demand.

Signs are rising that trade is already starting to deteriorate. Korea's technology exports slipped in July for the first time in more than two years with memory chips leading the falls. Semiconductor inventories piled up in June at the fastest pace in more than six years. Samsung and S.K. Heinex control roughly two-thirds of the global memory market, meaning they have the power to narrow the gap between supply and demand. Memory is loosely tied to other types of semiconductors built by firms such as TSM that provides chips in iPhones and invidia, whose graphics cards are used in everything from games to crypto mining and artificial intelligence, end quote. Sources are telling Bloomberg that for years, regulators have been hindered by banks' usage of messaging apps. like WhatsApp to skirt archiving rules. Around 10 banks might be fined around $2 billion in total because of this. Quote, investigators at the Securities and Exchange Commission and Commodities

Futures Trading Commission were repeatedly hindered by firms not archiving communications as required, according to people familiar with the matter. The watchdogs worried that missives on bankers' personal phones about cutting deals, trading, and courting clients were being completely lost and would ultimately make it harder to look for wrongdoing. At the SEC, separate probes revealed a troubling dynamic. Key conversations across finance were happening beyond the government's reach, according to one of the people. At the CFTC, similar concerns grew as officials probed weather banks were manipulating the interest rates swaps market, and they found that many communications were happening outside of official channels, people said. The scrutiny intensified at the SEC after Chair Gary Gensler took over in April

2021. After investigating J.P. Morgan Chase and Company over the lapses, the regulator opened an industry-wide sweep across Wall Street to figure out how many business-related communications were missing. The crackdown is now expected to result in about 10 banks, paying fines totaling around $2 billion with lenders from Goldman Sachs to Barclays, saying they expect comparable penalties to J.P. Morgan, which announced in December it would pay $200 million in penalties to the SEC and CFTC. That dwarfs the $15 million, Morgan Stanley agreed in 2006 after being accused of similar lapses. During a tense meeting over allegations that J.P. Morgan executives routinely shirked surveillance duties by tapping out work communications on personal messaging platforms and email,

SEC officials said previous fines shouldn't serve as guideposts for the penalty the bank would have to pay to settle the case. Enforcement director Gerbier Grewell said past punishments hadn't gotten banks to take record keeping seriously enough, according to people familiar with the conversation, end quote. Amazon has apparently begun telling merchants using fulfillment by Amazon to pay a fee of 35 cents per item sold in the U.S. and Canada beginning October 15th and lasting through January 22nd. Amazon is citing rising expenses, but this is the first time ever that they've jacked up prices around the key holiday sales season. Quoting CNBC, the holiday fee comes on top of existing

charges that sellers pay for using FBA services. Those costs vary depending on an item size, category, and weight. Amazon says it's implementing an added holiday surcharge for the first time as, quote, expenses are reaching new heights, making it harder for the company to absorb costs tied to the peak shopping season. Our selling partners are incredibly important to us, and this is not a decision we made lightly. Amazon said in the email, sellers fees have already been going up this year as Amazon manages through higher costs tied to inflation, supply chain constraints, and the war in Ukraine. Amazon's expenses also jumped due to rapid hiring and warehouses to meet a pandemic-driven surge in e-commerce demand. At a certain point, you can't keep absorbing all these costs and run a business that's economic.

CEO Andy Jassy told CNBC in an interview in April, end quote. And chain analysis says that so far this year through July, crypto transaction volumes are lagging behind 2021. Illicit volumes are down 15% year-over-year, and legitimate volumes are down to the 36% year over year. Quote, total scam revenue for 2022 currently sits at $1.6 billion, 65% lower than where it was through the end of July in 2021, and this decline appears linked to declining prices across different currencies. Since January 2022, scam revenue has fallen more or less in line with Bitcoin pricing. And as we see on the chart below, it's not just scam revenue falling. The cumulative number of individual transfers to scam so far in 2022 is the

lowest it's been in the past four years. Those numbers suggest that fewer people than ever are falling cryptocurrency scams. One reason for this could be that with asset prices falling cryptocurrency scams, which typically present themselves as passive crypto investing opportunities with enormous promise returns, are less enticing to potential victims. We also hypothesize that new inexperienced users who are more likely to fall for scams are less prevalent in the market now that prices are declining as opposed to when prices are rising, and they're drawn in by hype and the promise of quick returns. However, no area of cryptocurrency-based crime is bucking the 2022 downtrend of declining revenue like stolen funds. Through July 2022, $1.9 billion worth of cryptocurrency has been stolen in

hacks of services compared to just under $1.2 billion at the same point in 2021. Much of this can be attributed to the stunning rise of funds stolen from Defi Protocols, a trend that began in 2021. As we've covered previously, DeFi Protocols are uniquely vulnerable to. hacking as their open source code can be studied ad nauseum by cybercriminals looking for exploits, though this can also be helpful for security as it allows for auditing of the code. And it's possible that protocols incentives to reach the market and grow quickly lead to lapses in security best practices, end quote. This was surprising to me.

Quick. Which of the major tech platforms has been the biggest experimenter in the crypto space? Well, according to block data, it's alphabet, which invested around. $1.5 billion in blockchain companies between September 2021 and June 2022, the most by any public company, followed by BlackRock, with investments of $1.17 billion. Alphabet's bets have concentrated on four blockchain companies, including digital asset custody platform, fireblocks, Web 3 gaming company Dapper Labs, Bitcoin Infrastructure Tool Voltage, and venture capital company Digital Currency Group. This is in StarCob.

contrast to last year when Google diversified its much smaller $601.4 million funding effort across 17 blockchain-based companies, which again included Dapper Labs along with Alchemy, blockchain.com, cello, helium, and ripple. Google's increased investment into the blockchain industry is consistent with the other top 40 publicly traded companies with $6 billion in total being invested during this time compared to $1.9 billion between January 2021 to September 2021 and $506 million in all of 2020. The other big corporate investors include asset management company BlackRock, which invested $1.17 billion, investment banking corporation Morgan Stanley, investing $1.1 billion, and electronics company Samsung with investments totaling $979.79.2 million.

Like Google, Morgan Stanley and BlackRock adopted a more concentrated approach investing in only two to three companies during the period. However, Samsung was by far the most active investor having invested in 13 different companies, end quote. Finally today, forget robots writing code for you or drawing pictures for you. Google says it is using its most powerful large language model, known as Palm, to help robots from Alphabet X spinout company everyday robots understand complex human commands. So they're letting the bots loose on the bots, quoting Wired. late last week, Google research scientist Fay Jia sat in the center of a bright open-planned kitchen and typed a command into a laptop connected to a one-armed, wheeled robot resembling a large floor lamp.

I'm hungry, he wrote. The robot promptly zoomed over to a nearby countertop, gingerly picked up a bag of multi-grained chips with a large plastic pincer, and wheeled over to Zia to offer up a snack. The most impressive thing about that demonstration held in Google's Robotics Lab in Mountain View, California, was that no human coder had programmed the robot to understand what to do in response to Jia's command. Its control software had learned how to translate a spoken phrase into a sequence of physical actions using millions of pages of text scraped from the web. That means a person doesn't have to use specific pre-approved wording to issue commands, as can be necessary with virtual

assistants such as Alexa or Siri. Tell the robot, I'm parched, and it should try to find you something to drink. Tell it, whoops, I just spilled my drink, and it ought to come back with a sponge. Google's demo was a step toward the longstanding goal of creating robots capable of interacting with humans in complex environments. In the past few years, researchers have found that feeding huge amounts of text taken from books or the web into large machine learning models can yield programs with impressive language skills, including OpenAI's text generator GPT3. by digesting the many forms of writing online software can pick up the ability to summarize or answer questions about text, generate coherent articles on a given subject, or even hold cogent conversations.

The robot demonstrated was powered by the most powerful language model Google has announced so far, known as Palm. It is capable of many tricks, including explaining a natural language how it comes to a particular conclusion when answering a question. The same approach is used to generate a sequence of steps that the robot was. will execute to perform a given task. Researchers at Google worked with hardware from everyday robots, a company spun out of Google Parent Alphabet's X division dedicated to moonshot research projects to create new robot butlers. They created a new program that uses the text processing capabilities of Palm to translate a spoken phrase or command into a sequence of appropriate actions, such as open drawer or pickup chips, that the robot can perform. The robot's library of physical

actions was learned through a separate training process in which humans remotely controlled the robot to demonstrate how to do things like pick up objects. The robot had a limited set of tasks that it can perform within its environment, which helps prevent misunderstandings by the language model from becoming errant behavior, end quote. For now, I would say. Hey, Dali, too, animated GIF of Isaac Asimov nodding while stirring a cup of tea. We weren't able to squeeze in a kitchen table recording yesterday, but Chris Messina and his lovely partner were in town. And my wife and I had a great dinner with them in Williamsburg last night, as you might have seen on my Twitter feed. Chris and I were friendly before COVID times, but one of the best things about the last two years has been how Chris has become a truly close personal friend of mine.

Appreciate you, brother. And no, it's not worth doing a segment about Elon Musk tweeting about buying Manchester United. But geez, given what he's done to Twitter, if he took over that team, maybe that's the only way things could get more dysfunctional for them. Talk to you tomorrow.