Good Monday Morning

It’s August 21st. Friday is the deadline to add your name to the Facebook privacy class action settlement. You qualify if you were a U.S. Facebook user between 2007 and 2022. Official website.

Today’s Spotlight is 1,175 words — about 4 1/2 minutes to read.

3 Stories to Know

1. Revenge Porn Case: A Houston jury awarded over $1 billion in a revenge porn case, possibly setting a legal precedent. The decision underscores the issue’s potential damage. A 2017 study by Data & Society Research Institute found that 1 in 25 Americans has experienced nonconsensual image sharing.

2.Musk Throttles web: Twitter, now known as X, is slowing traffic to sites including the New York Times, Facebook, and Instagram, forcing users on X to wait an additional five seconds after clicking a link. The action targets companies that have previously drawn owner Elon Musk’s ire. Some throttling has stopped, but there are lingering concerns about Musk’s influence over user access to information.

3. Time & Weather: Google’s Contacts app now displays weather and time info for your contacts’ locations. The new feature will help communications across time zones and can be a good ice breaker too.

Spotlight on What to Know About AI

Recap of Part 1

Last week, we unraveled the basics of Artificial Intelligence (AI), the groundbreaking technology reshaping our lives. From voice assistants to personalized recommendations, AI is becoming an integral part of everyday experiences. But there’s more to the story!
 

Mystery Behind AI Outputs

AI might seem magical, but behind the scenes, it’s a complex data science.

Think of AI as a black box where data goes in and intelligent decisions come out. What happens inside? Algorithms like decision trees sort data into categories, while neural networks, akin to a web of interconnected brain cells, process information through layers, refining it into smart actions. Human experts often check these processes to ensure fairness and accuracy.

Then there’s generative AI like ChatGPT or Google Bard, a sophisticated output that functions like an advanced autocomplete system. Curious about how machines learn? Here’s an explanation made simple.

Ethics & Controversy

Ethical controversies surround AI development and use. The litany includes bias, accountability, and ownership of the output.

Popular video conferencing platform Zoom faced significant backlash this month when it revealed plans to train AI algorithms using customer calls. The company publicly canceled those plans days later amid horrific brand damage. The incident has led to a larger conversation about consent and transparency in AI development.

Taking a bold step, the New York Times blocked AI training on its content, signifying a turning point in how organizations address AI interactions. A week later, Microsoft joined them, a remarkable move for the company that has invested $10 billion in OpenAI and ChatGPT. Their moves reflects growing concerns about how AI algorithms might misinterpret or misuse journalistic or technical content, and it has spurred other media outlets to evaluate their own policies. 

Fair compensation for source material, one of the internet’s biggest bugaboos, is at the heart of many disputes.  

Google’s approach to AI has led to several debates about ethics and fair practice. One significant controversy is their use of online content for AI training, sparking concerns about copyrights. By leveraging publicly available information without explicit permission, questions arise about intellectual property rights and fair use.

The 2020 termination of renowned AI ethicist Timnit Gebru from Google ignited a firestorm of criticism. Gebru was a prominent advocate for diversity in technology and raised critical questions about bias in AI, specifically the large language models now at the heart of Google Bard.

Her dismissal exposed underlying tensions within the tech community about freedom of speech, research integrity, and the responsible development of AI. Together, these controversies represent the complex intersection between technology, law, and ethics, with potential wide-reaching ramifications.

Today

AI is no longer a futuristic concept — it’s here today, impacting how we live, work, and interact. The journey into AI’s world uncovers innovations, challenges, and ethical dilemmas. As AI continues to evolve, so does our understanding of this fascinating technology. 

Our Practical AI section below covers each week’s highlights and news in this explosive new field.

Practical AI

QuotableRight now, with 1,000 hours of therapy time, we can treat somewhere between 80 and 90 clients. Can you treat 200, 300, even 400 clients with the same amount of therapy hours?

— Stephen Freer, Chief Clinical Officer of Ieso that oversees 650 therapists who may use AI to help with case documentation.

Google, Universal Negotiate: Google is in talks with Universal Music Group to negotiate a licensing agreement for using Universal’s music and videos to train Google’s AI models. The negotiations mark a new approach in machine learning, using media to enhance understanding of music and visual content. This partnership could set a precedent for collaboration between tech and entertainment industries in AI development.  Artists, especially striking Hollywood writers and actors, are keenly aware of this issue.

Tool of the Week: This free infographic is one of the best I’ve seen to help guide people on using ChatGPT, Bard, or other AI chatbots.

Did That Really Happen — Maui Misinformation

Dangerous misinformation circulated falsely claiming that Maui residents accepting FEMA assistance could lose their homes or property to the federal government, a claim The Associated Press debunked.

Conspiracy theorists also falsely claimed that former President Barack Obama’s home was untouched by fires in Hawaii, stirring up conservative outrage, until others pointed out that Obama’s Hawaii home is on a different island.

Following Up — Another Abuse of Facial Recognition

We’ve told you repeatedly about law enforcement agencies misusing facial recognition. There’s news about Porcha Woodruff, a pregnant woman from Detroit, who was wrongfully arrested for robbery and carjacking after an automated facial recognition search. Despite being visibly pregnant, she was handcuffed, held for 11 hours, and had her iPhone seized as evidence. Woodruff’s is the sixth instance where a Black person has been falsely accused of a crime by police misusing facial recognition.  

Protip — Reverse Image Search

Learn how to perform a reverse image search with ease using this step-by-step guide that can help you discover how to find the original source of an image, debunk fake photos, and identify objects, people, or locations in pictures using Google, Bing, or TinEye.

Screening Room — Country Crock’s Legendary Campaigns Loves Moms

Science Fiction World — New Ocean Floor Ecosystem 

Scientists using robots have uncovered an ecosystem thriving beneath the ocean floor. This previously unseen world, located in Earth’s crust, hosts diverse microbes that play a crucial role in the planet’s cycle.

Coffee Break — Steve Ballmer’s Hysterical 1986 Ad Parody 

Steve Ballmer, the world’s 10th richest person, was about 30 years old and rallying the Microsoft troops around their new operating system called Windows, when he made this amazing commercial parody.

Sign of the Times

Good Monday Morning


It’s August 14th. Here are agencies accepting donations to help Maui recover.

Today’s Spotlight is 931 words — about 3 1/2 minutes to read.

3 Stories to Know

1. Meta’s Legal ChallengeMeta wants a new hearing following a ruling that its ad-targeting system may have facilitated discrimination. The ruling allowed plaintiffs, including civil rights groups, to pursue claims of race, gender, and other bias in advertising. 

2.Acoustic Attack on Keystrokes: A newly discovered acoustic attack can steal data by detecting keystrokes with an alarming 95% accuracy. Using built-in microphones on devices, the method listens to the sounds of typing, translating them into the actual keys pressed. Heads up: this is elite level hacking, not someone looking for your Facebook password.

3. TikTok’s U.S. Retail Push: TikTok is venturing into e-commerce in the U.S. by selling made-in-China goods, aiming to leverage its massive user base. The move marks a significant shift for the platform and aligns with parent company ByteDance’s broader e-commerce ambitions.

Spotlight on What to Know About AI

This week, Spotlight begins to demystify the intricate world of AI, unpacking the terminology, unraveling the mechanisms, and delving into the current debates.

Essential Definitions

Machine Learning (ML): A method where computers analyze data to predict outcomes, like Netflix’s recommendation system.

Large Language Model (LLM): A specialized form of ML trained on vast text data, enabling tools like chatbots to communicate naturally with people.

Generative AI: AI models creating content like the text or images in Dall-E’s or MidJourney’s art generation.

The Human Touch in AI Training

Human involvement in AI extends far beyond mere data processing. It’s not just a handful of experts either; thousands of people worldwide are engaged in the meticulous task of labeling data. This process involves identifying and categorizing elements within images, texts, and other forms of data.

For instance, labeling various plant species in photos teaches an AI to recognize similar patterns, while in other cases, humans might be transcribing and annotating spoken language. This extensive human interaction underlines the complexity of AI and is essential to its success.

The magnitude of this human involvement also brings challenges, such as ensuring quality and dealing with the sheer volume of data that requires annotation. Some projects might involve millions of individual pieces of data, each one needing precise labeling. While self-teaching algorithms exist, they concern scientists due to potential inaccuracies, emphasizing the irreplaceable role of human insight.

Machine Learning vs. True AI

ML predicts and analyzes data, forecasting weather patterns, whereas true AI (AGI) would interpret the weather’s impact on daily life. While ML is prevalent in tools like Google’s search engine, AGI is still a concept from science fiction, epitomized by HAL 9000 from “2001: A Space Odyssey” or Data from “Star Trek”. 

Coming Next Week

Next week, Spotlight takes you further into the realm of AI, exploring ethical dilemmas, corporate strategies, and the latest corporate events at Zoom, Google, and the New York Times  that are impacting the industry.

Practical AI

QuotableThese tools don’t work. They don’t do what they say they do. They’re not detectors of AI.”
— Deborah Weber-Wultf, a member of a multi-university research group evaluating 14 software programs that purportedly detect AI output.

NYC’s AI Bias Law: New York City’s new law targets AI bias in hiring, requiring audits for discrimination. This pioneering legislation aims to promote transparency and fairness in employment, while again, keeping humans in the mix.

Tool of the Week: Lifewire provides a walk-through demonstrating how to use Google Bard in Google Sheets.

Did That Really Happen — Heinz Ketchup’s Sugar Content

A viral social media claim stated that Heinz Tomato Ketchup contains a staggering amount of sugar. Snopes found the claim partially true; while Heinz ketchup does contain sugar, the amount is in line with industry standards and clearly labeled on the packaging. 

Following Up — Amazon’s Packaging Revolution

We told you last week that Amazon is retrenching. Now there’s news that Amazon may ditch unnecessary outer packaging for its Prime customers, aiming to reduce waste and save on costs. There are significant savings from reduced materials and shipping weights if it works. The long-term impact on both Amazon’s bottom line and the environment is yet to be determined.

Protip — Venmo Privacy Alert

Venmo’s social feed can inadvertently expose sensitive financial details. If you’ve ever used the platform, follow Brian Chen’s advice to review your privacy settings to ensure that your personal information remains confidential.’

Screening Room — Hyundai puts its family SUV in Grand Theft Auto

Science Fiction World — New Force of Nature? 

Researchers at Fermilab near Chicago are closing in on the potential discovery of a fifth fundamental force of nature, building on findings from 2021. Evidence from the behavior of sub-atomic particles called muons suggests they are being influenced by an unknown force. This discovery, part of the ‘g minus two (g-2)’ experiment, may lead to a revolution in physics.

 Coffee Break — Smarter Than A Scammer?

The Washington Post has another great interactive quiz for you to test your wits against digital scammers.

Sign of the Times

Good Monday Morning


It’s August 7th. News broke Sunday that U.S. scientists successfully repeated their fusion tests on July 30 with higher energy yields. 

Today’s Spotlight is 882 words — about 3 1/2 minutes to read.

3 Stories to Know

1. Hollywood to Cameo’s Rescue: Amid the ongoing strike, Hollywood actors are turning to Cameo for income, providing a much-needed boost for the previously embattled video greeting company. In July, layoffs had reduced Cameo’s head count to fewer than 50, as reported by The Information.

2.Prospect Medical Battling Cyberattack: Prospect Medical, an 18-hospital chain in five states, is battling an ongoing cyberattack. In response, they’ve taken their computer systems offline, suspending all nonemergency services, including elective surgeries and outpatient appointments. The FBI said it was working on the case with other law enforcement agencies.

3. Meta’s Preemptive News Block in Canada: In response to Canada’s Bill C-18, yet to take effect, Meta has blocked Canadian news content on Facebook and Instagram. The soon-to-be-enforced law requires tech giants to negotiate payments for hosting news content. Google also plans a similar move, echoing the companies’ 2021 Australian actions and earlier threats in California.

MidJourney 5.2, prompted by George Bounacos

Spotlight on Amazon Retrenching 

Amazon Refocuses Amid Challenges: After a pandemic-induced surge and subsequent hit to its stock price and reputation, Amazon has been leaning harder than ever on its core cloud services and e-commerce businesses, experiencing a 65% increase in its market cap this year. Wall Street has welcomed the company’s job cuts of over 27,000, allowing room for key initiatives:

1. Tech Expansion: Amazon channeled $100M into its emerging generative AI program, mirroring its successful transformation of cost centers like logistics into profit engines. With ad revenue up 20% plus quarterly, hitting $10.6B in Q2, AI is next in line. Also, in a competitive move, Amazon’s AI-powered coding assistant, CodeWhisperer, is now free, rivaling Microsoft’s $10 monthly charge.

2. Grocery Consolidation: Amid store closures, Amazon is testing a unified checkout process for Amazon Fresh and Whole Foods, bypassing certain ingredient restrictions long in place at the upscale grocer In another move, home delivery service, currently exclusive to Prime members, is being trialed for non-Prime members.

3. Health care VentureAmazon Clinics now offer nationwide video visits, a recent expansion. Operating independently of insurance, they offer potentially insured prescriptions, including through Amazon Pharmacy. Recall Amazon’s earlier $3.9 billion acquisition of OneMedical, bringing in 200 physician locations across major markets.

Hurdles remain despite these advances. Amazon’s purchase of Roomba creator iRobot suffered a privacy blow with leaked videos from Roomba development units. Critics spotlight Amazon’s pollution footprint, especially shipping materials and fuel for transporting returns. Despite efforts, including a UPS deal for no-cost returns and phasing out plastic padded shipping bags, critics remain unsatisfied.

Practical AI

QuotableWe were surprised by how much the N-word was there.” — Gaming executive Justin Liebregts whose studio used AI software to monitor player conversations for toxicity.

Noteworthy: A multi-university study of 14 generative AI tools concluded that it remains relatively easy to fool software designed to tell if content was AI or human-created.

Tool of the Week: Meta released three new text-to-music generators. You can try the most basic and make a 12 second audio clip free.

Did That Really Happen — Investor Arrested

Fabian Marta, one of the 6,000-plus investors in the theatrical release of the movie “Sound of Freedom,” was arrested in St. Louis on July 23 and charged with felony child kidnapping. Despite social media rumors, Marta was not a producer for the film, but contributed to its $5 million theater budget.

Following Up — Chromebook Obsolescence

Back-to-School Warning:Last week, we reported on third-party sellers offering soon-to-be unsupported Google Chromebooks on platforms like Amazon and Walmart. This week, a report from U.S. PIRG called the situation “absurd” and revealed eight ChromeOS devices with an impending “death date” are currently available for purchase.

Protip — Personal Info Help

Google has rolled out a feature to aid in locating search results containing your personal information and assist in their removal. The service also tracks the status of past removal requests and sets alerts for when new copies are detected online. More details available here.

Screening Room — Debinha Dazzles for Nike Football

Science Fiction World — AI-Powered Medical Milestone 

Surgeons have successfully implanted AI-enabled microchips in a quadriplegic man’s brain, leading to the restoration of some feeling in his hand. This breakthrough included five chips connected to electrodes and sensors on the patient’s arm and hand.

Coffee Break — Most Polluted Cities

The American Lung Association has an updated State of the Air report that allows you to find the cleanest and most polluted places–and see how your city stacks up. 

Sign of the Times