Tech Brew Ride Home - Fri. 09/08 – An AI To Talk To The Animals?

Episode Date: September 8, 2023

You really need to update your Apple devices. Today. Here come the AI generated misinformation campaigns. Microsoft will defend you from lawsuits if you use their AI. And in the Weekend Longreads Suggestions, might AI finally give us a real life Doctor Dolittle? Sponsors: CrucibleMoments.com Nutrafol.com/men promocode: ridehome Links: Apple discloses zero-days linked to NSO Group spyware (The Record) Ex-FTX Executive Ryan Salame Could Forfeit $1.5B as Part of Guilty Plea (CoinDesk) China turns to AI in hopes of creating viral online propaganda, Microsoft researchers say (Cyberscoop) Microsoft Says It Will Protect Customers from AI Copyright Lawsuits (Bloomberg) Exclusive: ChatGPT traffic slips again for third month in a row (Reuters) Weekend Longreads Suggestions: The Decomposition of Rotten Tomatoes (Vulture) What OpenAI Really Wants (Wired) How artificial intelligence could help us talk to animals (ScienceNewsExplores) 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 TechMemebride home for Friday, September 8th, 2020. I'm Brian McCullough today. You really need to update your Apple devices right now.

Here come the AI-generated misinformation campaigns. Microsoft will defend you from lawsuits if you use their AI. And in the weekend, long-rate suggestions, might AI finally give us a real-life Dr. Doolittle? Here's what you miss today in the world of tech. Public service announcement. Update all of your Apple devices, Toot Suite. because Apple has released MacOS, iOS, iOS, iPodOS, and watchOS updates to address two zero-day flaws that Citizen Lab says were already being used to deliver that NSO Group's Pegasus Spyware,

quoting the record. Cybersecurity researchers at the Citizen Lab at the University of Toronto said that all users of Apple devices should update their operating systems immediately to fix the bugs. Last week, while checking the device of an individual employed by a Washington, D.C. based civil society, organization with international offices, Citizen Lab found an actively exploited zero-click vulnerability being used to deliver NSO Group's Pegasus Mercenary Spyware, the researchers said. The exploit chain was capable of compromising iPhones running the latest version of iOS 16.6, without any interactions from the victim, the researcher said. One bug tracked as CVE 2023-41064, allowed devices including some iPhones, iPads,

Macs and Apple Watches, to become vulnerable to attack when processing a maliciously crafted image, Apple said. It affects the image I.O. Framework specifically. The other vulnerability CVE 2023-41061, similarly creates security issues if a device is sent a maliciously crafted attachment. That bug arose in the company's wallet function. In both cases, Apple said it was aware of a report that this issue may have been actively exploited. Apple declined to comment more about the bugs. Citizen Lab said it had immediately disclosed our findings to Apple and assisted in their investigation. The software updates apply to MacOS Ventura, iOS, iPodOS, and watchOS. The patches were made available as part of regular updates to those products. They were not labeled as rapid security responses.

The term Apple uses for bug fixes issued urgently between full OS updates. With the disclosure of these two vulnerabilities, the company has now patched 13-0 days in 2023. Since it was first developed in 2011, Pegasus has been used across the globe, often by governments spying on their citizens. It has been deployed to target assassinated Saudi journalist Jamal Khashoggi, members of the Catalan independence movement and human rights investigators in Mexico. In recent years, regulators have attempted to prevent its spread, with the European Parliament urging EU member nations to ban it, and U.S. President Joe Biden signed an executive order earlier this year,

blocking the use of commercial spyware by the government, end quote. former FTX Digital Markets co-CEO Ryan Salami has pled guilty to U.S. campaign finance and money transmitting charges and will forfeit more than $1.5 billion as part of the deal, according Coin Desk. Salami, who was co-CEO of FTX's Bahamas entity FTX Digital Markets, pleaded guilty to conspiracy to make unlawful contributions and to fraud the Federal Election Commission and conspiracy to operate an unlicensed money transferring business. I made political contributions in my name that were funded by transfers from an Alameda subsidiary, Salami told Judge Lewis Kaplan, who is also overseeing Sam Bankman-Freed's trial as he entered his guilty plea.

The transfers were categorized as loans, Salami said, but, quote, it was understood that they would not be repaid. The donations, according to Salami, were for the benefit of initiatives introduced by others, but supported by Sam Bankman-Fried, end quote. As part of his plea agreement with the government, Salami has been ordered to forfeit more than $1.5 billion. He agreed to forfeit $6 million before his sentencing expected in March of next year. To help cover this amount, Salami has already agreed to give the government a 2021 Porsche automobile and multiple properties, including two Massachusetts homes and ownership of the East Rood Farm Corporation an entity Salami owns. Should Salami pay the $6 million and turn over the various properties

by the set deadlines referred to as the substitute assets, he will be off the hook for the amount, a DOJ filing said. Upon receipt of the payment, the U.S. shall accept the payment and substitute assets in full satisfaction of the money judgment, the filing said. The DOJ could still pursue the full $1.5 billion if it turns out that Salami lied in a financial affidavit or if he otherwise interferes with the government's attempt to deal with the assets, the document said. Additionally, Salami was ordered to pay more than $5.5 million in restitution to FTX debtors. According to a different DOJ document, the $1.5 billion salami will forfeit represents property involved in the unlicensed money transmitter charge, end quote. I guess we should expect this to be only the beginning

of something like this. Microsoft says Chinese state-affiliated hackers are using AI to generate content designed to, quote, go viral across social networks in the U.S. and other democracies, quoting CyberScoop. We have observed China-affiliated actors leveraging AI-generated visual media in a broad campaign that largely focuses on politically divisive topics such as gun violence and denigrating U.S. political figures and symbols, the researchers said. The visuals generated during the campaign are more, quote, eye-catching than the awkward visuals previously deployed in Chinese operations, the researchers noted in a report and blog post published Thursday morning. The activity is part of Chinese information operations increasing

success at engaging target audiences around the world, which includes, quote, China State-affiliated multilingual social media influencer initiative that has, quote, successfully engaged target audiences in at least 40 languages and grown its audience to over 103 million, the researchers said in the report. Chinese state-sponsored propaganda is pushed by a network of more than 230 state media employees and affiliates who masquerade as independent social media influencers across all major Western social media platforms, the researchers said. Part of this activity includes social media personas operated by real people,

that employ fictitious or stolen identities that conceal connections with the Chinese government and share artificially generated content. This relatively high-quality visual content has already drawn higher levels of engagement from authentic social media users, the researchers said. Simultaneously, Chinese-related cyber espionage operations continue apace, including the operation revealed in May, where hackers used stolen Microsoft authentication keys to steal U.S. cabinet officials' emails ahead of strategic diplomatic meetings between U.S. and Chinese officials. Microsoft said, in a separate blog post Wednesday that the Chinese Hacking Group in that case, tracked as Storm 0558, compromised a Microsoft engineer's corporate account in April 2021

to gain access to the keys. Microsoft also revealed in May that Chinese-linked hacking operations targeted critical infrastructure entities in the U.S. and Guam, possibly as part of efforts to lay the groundwork for disrupting communications between the U.S. and Asia in the event of a crisis. That operation was linked to a group tracked as Volt Typhoon, but at least two other distinct Chinese link hacking groups continue to target the U.S. defense industrial base broadly, according to the report. China has displayed a, quote, particular focus on the South China Sea region. The researchers said in Thursday's report in activity that, quote, signals attempts to gain competitive advantages for China's foreign relations and strategic military aims, end quote. Microsoft has committed to defending

its customers, at least customers of its AI co-pilots products from copyright infringement lawsuits. as they've, quote, use the guardrails and content filters effectively. Quoting Bloomberg. Microsoft also pledged to pay related fines or settlements and said it has taken steps to ensure its co-pilots respect copyright. We believe in standing behind our customers when they use our products, Microsoft said. We are charging our commercial customers for our co-pilots, and if their use creates legal issues, we should make this our problem rather than our customer's problem, end quote. Generative AI applications scoop up existing content such as art, articles, and programming code, and use it to generate new material that can simplify

or automate a range of tasks. Microsoft is baking the technology developed with partner OpenAI into many of its biggest products, including Office and Windows, potentially putting customers in legal jeopardy. It's not the first time Microsoft has deployed a legal shield to keep customers loyal. In the 2000s, the company offered indemnification to partners and later customers using or reselling its software, a bid to differentiate Microsoft from Linux and other open source software makers. In 2017, Microsoft by then a seller of open source software itself offered to protect customers of its Azure cloud products from legal claims. The company in June announced a program to help customers ensure the AI programs they run on Microsoft platforms meet global

laws and regulations. Earlier this year, Adobe also offered subscribers of its AI tools legal protection against copyright infringement, end quote. to similar web, usage of ChatGPT continues to decline. ChatGPT's desktop and mobile website visits fell 3.2% month over month to $1.43 billion in August, following 10% drops in each of the previous two months. But there might be a reason for this, quoting Reuters. The amount of time visitors spent on the website has also been declining monthly since March, from an average of 8.7 minutes on site to seven minutes on site in August. But August worldwide, you unique visitors ticked up to 180 and a half million users from 180 million. School coming back

into session in September may help chat GPT's traffic and usage, and some schools have begun to embrace it. U.S. chat GPT traffic in August rose slightly in concert with American schools being back in session. Students seeking homework help appears to be part of the story. The percentage of younger users of the website dropped over the summer and is now starting to bounce back, said David F. Carr of Similar Web, who regularly tracks chat G. P.T and its competitors. Before Meta's threads launch, chat GBT was the fastest growing consumer application ever and is now one of the top 30 websites in the world, end quote. Time for the week and long way suggestions. We recently had a piece about whether or not

the quality of the website wirecutter had gone downhill. Well, this piece from Vulture asked if rotten tomatoes is crap now too, largely due to people gaming it. Quote, another problem, and where the trickery often begins, is that Rotten Tomato Scores are posted after a movie receives only a handful of reviews, sometimes as few as five, even if those reviews may be an unrepresentative sample. This is sort of like a cable news network declaring an election night winner after a single county reports its results. But studios see it as a feature, since with a little elbow grease, they can sometimes fool people into believing a movie is better than it is. Here's how. When a studio is prepping the release of a new title, it will screen the film for critics in advance.

It's a film publicist job to organize these screenings and invite the writers they think will respond most positively. Then that publicist will set the movie's review embargo in part so that its initial tomato meter is as high as possible at the moment when it can have maximal benefits for word of mouth and early ticket sales. Granted, that is not rocket science or even particularly new, but the strategy can be surprisingly effective on tent pole releases for which studios can leverage the growing universe of fan-run websites whose critics are generally more admiring of comic book movies than those who write for mainstream outlets. No offense to comicbook movie.com. For example, in February, the tomato meter score for Ant Man and the Wasp, Quantummania debuted at 79% based on its first batch of reviews. Days later, after more critics had weighed in, its rating sank into the 40s. But the gambit may have worked.

Quantummania had the best opening weekend of any movie in the Ant Man series at $106 million. In its second weekend, with its rottenness more firmly established, the film's grosses slid 69% the steepest drop-off in Marvel history, end quote. Then everybody's been talking about this big profile of open AI in Wired by the great friend of the show Stephen Levy. Quote, for Altman and his company, Chat GPT and GPD4 are merely stepping stones along the way to achieving a simple and seismic mission. One, these technologists may as well have branded on their flesh.

That mission is to build artificial general intelligence, a concept that's so far been grounded more in science fiction than science, and to make it safe for humanity. The people who work at OpenAI are fanatical in their pursuit of that goal, though, as any number of conversations in the office cafe will confirm, the build AGI bit of the mission seems to offer more raw excitement to its researchers than the make it safe bit. These are people who do not shy from casually using the term superintelligence. They assume that AI's trajectory will surpass whatever peak biology, can attain. The company's financial documents even stipulate a kind of exit contingency for when AI wipes away our whole economic system. It's not fair to call OpenAI a cult, but when I asked several of the company's top brass if someone could comfortably work there if they didn't believe AGI was truly coming and that its arrival would mark one of the greatest moments in human history, most executives didn't think so. Why would a non-believer want to work here, they wondered? The assumption is that the workforce now at approximately 500, though it might have grown since you began reading this paragraph, has self-selected to include only the faithful. At the very least,

as Altman puts it, once you get hired, it seems inevitable that you'll be drawn into the spell, end quote. And finally today, think about what it is that generative AI actually does. It finds patterns, right? It parses language. So what if generative AI is on the verge of fulfilling one of humankind's oldest witches? The ability to talk to and understand animals, quoting Science News, Explorers. is a linguist at the University of California, Berkeley. He got the chance last summer to observe sperm whales in their wild Caribbean habitat off the coast of the island nation of Dominica. With him, were marine biologists and roboticists. There were also cryptographers and experts in other

fields. All have been working together to listen to sperm whales and figure out what they might be saying. They call this project SETI. That's short for Cetation Translation Initiative, because sperm whales are a type of cetacean. To get started, Project SETI has three three listening stations. Each one is a cable hanging deep in the water from a buoy at the surface. Along the cable, several dozen underwater microphones record whale sounds. From the air, drones record video and sounds. Soft, fish-like robots do the same underwater. Suction cup tags on the whales capture even more data. But just collecting all these data isn't enough. The team needs some way to make sense of it all. That's where artificial intelligence or AI comes in. A type of AI known as machine learning

can sift through vast amounts of data to find patterns. Thanks to machine learning, you can open an app and use it to help you talk to someone who speaks Japanese or French or Hindi. One day, the same tech might translate sperm whale clicks. Project SETI's team is not the only group turning to AI for help deciphering animal talk. Researchers have trained AI models to sort through the sounds of prairie dogs, dolphins, naked mole rats, and many other creatures. Could their efforts crack the codes of animal communication? End quote. San Francisco. I am coming to you. A month from today, I'll be attending the first ever AI Engineer Summit.

It's being organized by, among other folks, friend of the show, Sean S-W-Y-X-W-Wang. You can find out more at the website, AI.Engineer. I will be there October 8th through the 10th, so Chris and I will probably try to organize another podcast listener meetup. If folks are interested, more details on that when we have it. But please check out AI. engineer, the first ever conference for this new discipline, this new career field should be cool to attend. Talk to you on Monday.