Tech Brew Ride Home - Thu. 06/08 – The Gloves Come Off In The Binance Case

Episode Date: June 8, 2023

So, the Binance case is already getting testy with some wild accusations flying in both directions. Twitch walks back some ad guidelines after only two days. WhatsApp is adding channels. Prime Video is probably about to add ads. Adobe will protect you from lawsuits if you use their AI tools. And how AI might have just extended Moore’s Law a bit. Sponsors: EarnIn App (type in techmeme under PODCAST when you sign up) Links: Binance lawyers allege SEC Chair Gensler offered to serve as advisor to crypto company in 2019 (CNBC) SEC says Binance and US affiliate redirected billions in customer assets to Zhao's funds (The Block) Twitch walks back controversial ad rules policy (The Verge) WhatsApp’s new Channels feature brings social media to your messaging app (The Verge) Amazon Plans Ad Tier for Prime Video Streaming Service (WSJ) Adobe is so confident its Firefly generative AI won’t breach copyright that it’ll cover your legal bills (Fast Company) Google claims that Bard is improving at math and programming (TechCrunch) Google DeepMind’s game-playing AI just found another way to make code faster (MIT Technology Review) 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 Thursday, June 8th, 2023. I'm Brian McCullough today. So the Binance case is already getting testy with some wild accusations flying in both directions.

Twitch walks back some ad guidelines after only two days. WhatsApp is adding channels. Prime Video is probably about to add ads. Adobe will protect you from lawsuits if you use their AI tools. And how AI might have just extended Moore's Law a bit. Here's what you missed today in the world of tech. Yikes. The gloves are coming off already in the Binance legal fight. According to court filings,

Binance's lawyers are alleging SEC Chair Gary Gensler offered to be an advisor to Binance back in March of 2019 when he was still teaching at MIT Sloan, quoting CNBC. Documents filed by the SEC on Wednesday indicate that attorneys from Gibson, Dunn, and Latham and Watkins to have Binance's law firms alleged that Gensler offered to serve as an advisor to the crypto exchange in several March 2019 conversations with Binance executives and founder Chang Peng Zhao. He eventually met Zhao in Japan for lunch later that month, the filing claims. At the time, Gensler was teaching at Massachusetts Institute of Technology Sloan School of Management. He was appointed head of the SEC in 2021 by President Biden, and over the past year, has come down hard on the crypto industry suing numerous companies for allegedly selling unregistered

securities. Before Gensler started going after Binance, he was trying to cozy up to the company, the lawyers say. The Wall Street Journal previously reported on Gensler's and Binance's relationship, citing internal Binance messages and a person close to the SEC chair. Both suggested that Binance approached Gensler. In the latest filing, the Gibson and Latham attorneys say that Zhao continued to stay in touch with Gensler after the March meeting. And at the future SEC chair's request, Zhao sat down for an interview with Gensler as part of a cryptocurrency course. was teaching at MIT. Later in 2019, the letter said Gensler was slated to testify before the House Financial Services Committee, and he sent Zhao a copy of his intended testimony ahead of the hearing.

In July of that year, Gensler testified before the House over Facebook's proposed and later canceled cryptocurrency, Libra, and its planned Calibro wallet. I do not advise any financial technology, blockchain, or other companies, nor do I own any cryptocurrencies. Gensler's prepared testimony read, end quote. Meanwhile, other court documents show the SEC, accusing Binance and Binance.U.S., of redirecting more than $12 billion in customer assets to entities controlled by CZ between 2019 and 2021, quoting the block. The fresh SEC allegations come as supporting evidence in a request by the agency to a federal judge to place a temporary freeze on assets beyond customer redemptions at Binance.us. The SEC also wants a judge to freeze Binance and Zhao's assets, but has only filed a temporary

restraining order on Binance.U.S.'s' assets. Lawyers for the SEC alleged that most of the assets were funneled to Merit Peak. The agency quotes a statement by Merit Peak as describing itself as a, "...proprietary trading firm with Zhao's self-made wealth from the digital assets business," end quote. Those funds consisted in significant part of Binance Platform's customers' assets, including those of Binance.coms platform customers and other sources. the SEC alleges. In total, the agency claims over $12 billion of the $22 billion, merit peak received between 2019 and 2021 came from Binance and Binance.U.S. Customer assets. Per the SEC's court statement, the $11 billion from Binance customer assets was funneled

through another entity called Key Vision Development Limited, end quote. We didn't get a chance to talk about this controversy, but Twitch is already walking back new branded content guidelines it rolled out just a few days ago, calling them bad for streamers and the company after a swift backlash from creators, which again was so swift, we didn't get a chance to even talk about it. Here's how it all went down according to the verge. Quote, on Tuesday, Twitch released new rules concerning the way streamers could display ads on the platform. The rules prohibited burned in video display and audio ads, the first two of which were popular and common formats used throughout Twitch. Twitch apparently did not discuss the new

rules with ambassadors or streamers beforehand, and many were furious about the new policies. Twitch content creators took to social media to decry the changes. OTP, a network of popular high-value streamers like Asman's Gold released an open letter to Twitch telling it, quote, the once unique and admirable vision of a creator-first platform now feels like a fading and distant dream. Charity streamers were upset and fearful, believing the new rules would impact their ability to raise money. It was the same with e-sports creators, as the new rules would have made it more difficult for the already struggling e-sports industry to monetize its broadcasts. Twitch apologized for the rollout, explaining that it would rewrite the rules for greater clarity. Now it seems that rewrite has turned into a full rescinding of the rules totally. Twitch has updated the page outlining its ads policy with the section related to what kinds of ads are prohibited or allowed, completely removed.

Here's an archive version with the old rules and the new updated page. The new rules would have been potentially devastating for creators, charities, e-sports, broadcasts, and brands. Now, what seemed like another attempt to take a portion of streamer earnings has backfired, end quote. This is potentially pretty big. WhatsApp is announcing channels, a one-to-many broadcast tool that is what they're calling a private way to follow what matters. They also plan to add payment services in the future, quoting the verge. The newest feature in WhatsApp brings a very different kind of messaging to the world's most popular chat app. It's called channels, and it's designed specifically for one-to-many broadcasts rather than conversational. But what is a channel really? It's a Twitter feed, minus all the metrics and reply guys. WhatsApp has

clearly noticed all the government's transit agencies, brands, and others looking for a new and non-Twitter place to share their most important updates and sees channels as a drop-in replacement. Channels is also a creator tool to some extent, a place for those with an audience to, quote, send text, photos, videos, stickers, and polls according to WhatsApp's launch blog post. The company has plans to build a payment and other monetization services into channels too. You'll be able to find channels by searching for them in WhatsApp or by browsing in a newly created directory and see their most recent updates in the status section of the app. WhatsApp says privacy is a key part of the experience, which is why channel admin information

isn't shared, and the app only stores 30 days' worth of a channel's history. Admins can even block screenshots and forwards making sure that what's in the channel stays in the channel. Channels aren't end-to-end encrypted, though. They're treated more like your messages with businesses, which are also not totally private. But WhatsApp does say it's thinking about ways to encrypt some channels over time. This is mostly a really obvious feature for WhatsApp to add. Telegram has a similar feature, also called channels, and also meant for one to many

broadcasts has had it for years. Instagram has a similar feature, too, called broadcast channels, and it actually makes sense to bring this kind of information into WhatsApp. Getting air quality updates and train statuses feels more natural in a messaging app than intermingled with everything else on Twitter. But if you zoom out a bit, WhatsApp is quickly turning into something. other than just a messaging app. Just in the past few months, the company has made it possible to use one account on multiple phones, has been working on a private newsletter tool and a new username system,

added polls and shopping, and a bunch of other Facebook-y things to the platform, revamped its status system, improved its group chats, and much more. Channels is just the latest way WhatsApp is trying to bring social media to messaging, end quote. Sources are telling the journal that Amazon plans to launch an ad-supported prime video tier and is talking with WBD and Paramount about adding their ad-based tiers to prime video channels. Quote, advertising has been an area of continued growth for Amazon despite macroeconomic challenges. The company's ad revenue was $9.5 billion in the first quarter, up 21% year over year. The company is the third biggest player in terms of digital ad revenue in the U.S.

after Google and meta, according to insider intelligence. Advertisers say they are eager to have Amazon offer an ad tier for prime video service, which would follow similar moves by other streaming platforms, including Netflix and Disney, Specifically, ad buyers say they want more access to premium movies and programs that have remained largely ad-free, content that often garners more buzz. Amazon has been making moves to bring more ad-supported programming to Prime Video. It's sports coverage comes with ads, and some shows have product placement advertising. Prime Video also offers access to content from FreeVee, Amazon's free ad-supported video service. Separately, the e-commerce giant is having discussions with Warner

Brothers Discovery and Paramount Global about adding the ad-based tiers of their streaming services through Prime Video Channels, according to people familiar with the situation. Through Prime Video Channels, users can sign up for streaming services, including the ad-free versions of Max and Paramount Plus, and view through the app, end quote. Adobe has begun rolling out its generative AI Firefly tool to businesses and plans to offer what it is calling full indemnification against copyright lawsuits over its output, quoting Fast Company. Anything created using Firefly's text-to-image generation tool will be fully indemnified by the company, quote, as a proof point that we stand behind the commercial

safety and readiness of these features, Adobe says. That's important because of the challenges around the legal status of generative AI tools and their outputs. The standards around generative AI and copyright have not yet been settled legally, which is causing companies to hold off using generative AI in their business operations. This decision, Adobe hopes, provides clarity. The Firefly model is trained on stock images for which Adobe already holds the rights, as well as on openly licensed content, for example, Creative Commons images and public domain content. Adobe has actually offered indemnification for quite some time against the use of its own products, and in particular for stock images, a representative for the company says, noting that this is

an extension of that practice. The representative declined to answer whether the indemnity means that anyone who believes their copyright has been infringed by Firefly should sue Adobe rather than the person who used Firefly, noting instead, it's a guarantee against litiganties. litigation, the consequences of litigation, end quote. The offer will be available only to enterprise customers, and Adobe declined to share how much the company had set aside for a legal fund to fight any lawsuits should they arrive, end quote. Google has updated how its Bard tool handles math, coding questions, and string manipulation via what it calls implicit code execution, which lets the chatbot run code in the background, coding tech crunch. As the blog post explains large language models such as

Bard are essentially prediction engines. When given a prompt, they generate a response by anticipating what words are likely to come next in a sentence. That makes them exceptionally good email and essay writers, but somewhat error-prone software developers. But wait, you might say, what about code generating models like GitHub's co-pilot and Amazon's Code Whisperer? Well, those aren't general purpose. Unlike Bard and rivals along the lines of chat GPT, which were trained using a vast range of text samples from the web, e-books and other resources, code pilot and Code Whisperer, and comparable code generating models were trained and fine-tuned almost exclusively on code samples. Motivated to address the coding and mathematics shortcomings in general LLMs, Google developed

implicit code execution, which allows Bard to write and execute its own code. The latest version of Bard identifies prompts that might benefit from logical code, writes the code, under the hood, tests it, and uses the result to generate an ostensibly more accurate response. Based on internal benchmarking, Google says that the new Bard's responses to computation-based word and math problems were improved by 30% compared to the previous Bard release. Of course, we'll have to see whether those claims stand up outside of testing. When Google launched Bard earlier this year, it didn't compare that favorably to the likes of Bing Chat and Chat CheapyT. Instead, the rollout was a bit of a disaster, with a Google ad featuring a wrong answer by Bard, briefly tanking the company's stock 8%.

Reportedly, several Google employees who tested Bard prior to its release raised serious concerns to the search giant, with one person calling it a pathological liar and another deeming it worse than useless. With implicit code generation and other enhancements like support for new languages, multimodal queries, and image generation, Google is responding to criticism and attempting to turn the situation around, end quote. And finally today, Deep Mines Alpha-Dev and AI for discovering enhanced algorithms devised new C++ sorting algorithms that are up to 70% faster and have subsequently been added to that language. Now, this might sound dull, but actually it points away to something important.

could AI extend the life of Moore's Law? Quoting MIT Technology Review. Last year, the company used a version of its game-playing AI Alpha-Zero to find new ways to speed up the calculation of a crucial piece of math at the heart of many different kinds of code, beating a 50-year-old record. Now it has pulled the same trick again twice. Using a new version of Alpha Zero called Alpha-Dev, the UK-based firm, recently renamed Google Deep Mind after a merge with its sister company's AI Lab in April, has discovered a way to sort

items in a list up to 70% faster than the best existing method. It has also found a way to speed up a key algorithm used in cryptography by 30%. These algorithms are among the most common building blocks in software. Small speedups can make a huge difference, cutting costs and saving energy. Moore's Law is coming to an end where chips are approaching their fundamental physical limits, says Daniel Mankowitz, a research scientist at Google Deep Mind, we need to find new and innovative ways of optimizing computing, end quote. DeepMind published its results in nature today, but the techniques that AlphaDev discovered are already being used by millions of software developers. In January 2022, DeepMind submitted its new

sorting algorithms to the organization that manages C++, one of the most popular programming languages in the world, and after two months of rigorous independent vetting, AlphaDev's algorithms were added to the language. This was the first change to C++'s soaring algorithms in more than a decade, and the first update ever to involve an algorithm discovered using AI. DeepMind added its other new algorithms to Abysseal an open source collection of pre-written C++ algorithms that can be used by anybody coding with C++. These cryptography algorithms compute numbers called hashes that can be used as unique IDs for any kind of data.

DeepMind estimates that its new algorithms are now being used trillions of times a day. Alpha-Dev could not beat the best human version of the algorithm for sorting a list of four items, which takes 28 instructions, but it beat the best human version for five items. cutting the number of instructions down from 46 to 42. That amounts to a significant speed up. The existing C++ algorithm for sorting a list of five items took around 6.91 nanoseconds on a typical Intel Skylake chip. Alpha devs took 2.01 nanoseconds around 70% faster. Scientists would like to see a more exhaustive comparison with the best human-divized approaches, especially for longer algorithms. DeepMind says that that's part of the plan.

Mankowitz wants to compare Alpha-Dev with the best human-devs methods getting the AI to build on human intuition rather than starting from scratch. After all, there may be more speed-ups to be found. For a human to do this, it requires significant expertise and a huge amount of hours, maybe days, maybe weeks, to look through these programs and identify improvements, says Mankowitz. As a result, it hasn't been attempted before, end quote. So I'm pretty sure you've seen those pictures from New York City yesterday.

about 3 p.m. Eastern yesterday it was just wild, just orange and dark. The pictures kind of don't do justice to how eerie it was to peek your head outside. Today I can report that it is bright and sunny with just a slight orange tint, but those air quality indexes all report it being still unhealthy, so we're all still inside again, running our air purifiers. Talk to you tomorrow.