Ideas that challenge / comfort / inspire
Stephen Goforth
Prompt Injection - Like prompt engineering (where a user is good at writing AI prompts), but, in this case, with the goal of working around AI to produce harmful content. Hackers use carefully crafted prompts or text-based instructions to manipulate generative AI systems into sharing sensitive information or perform unintended actions by making the model ignore previous instructions.
More AI definitions here
Many companies have an unspoken command floating through the halls: "Bow to the organization above the rest!" There is a constant re-evaluation as to whether someone is playing their loyal role for the tribe. Talk of employees taking time for family and self-care is just that—talk. In practice, the expectation is that everyone will constantly genuflect toward the hierarchical (and often paternalistic) structure.
Someone who drinks to excess, yells at coworkers, holds racist views, drives away competent employees, has materialistic goals, and so forth will be tolerated, even rewarded, as long as their allegiance is true, helped by bringing in dollars or playing some other role that helps to perpetuate the organization.
On the other hand, someone with none of those vices might be cast aside if they are deemed not adequately sacrificing themselves on the altar of the organizational machinery.
In the children's book "Hope for the Flowers," Trina Paulus tells the story of caterpillars who form a tower with their bodies. They climb over each other in an attempt to reach the top. What reward waits for them? Nothing. Nothing at all. The struggle to rise only serves to stop them from cocooning and becoming the butterflies they were meant to me.
Stephen Goforth
Publishers are selling papers to train AIs — and making millions of dollars – Nature
Can AI fix the peer review bottleneck? Top journals weigh in - News-Medical.Net
Can Google Scholar survive the AI revolution? - Nature
Rethinking Reviewer Fatigue AI’s place in a broader discussion - EON
Writing assistant, workhorse, or accelerator? How academics are using GenAI – London School of Economics
Can AI review the scientific literature — and figure out what it all means? – Nature
The great AI witch hunt: Reviewers’ perception and (Mis)conception of generative AI in research writing – Science Direct
Can AI be used to assess research quality? – Nature
Scholarly Publishing World Slow to Embrace Generative AI – Inside Higher Education
National Academies President on How to Use Generative AI Responsibly in Scientific Research – JAMA
Study of ChatGPT citations makes dismal reading for publishers – Tech Crunch
Experts fail to reliably detect AI-generated histological data – Nature
AI-authored abstracts ‘more authentic’ than human-written ones – Times Higher Ed
AI’s computing gap: academics lack access to powerful chips needed for research – Nature
AI-generated survey responses could make research less accurate – and a lot less interesting – Stanford
The sheriff’s department in a country just southwest of Nashville has deployed AI-powered dashcams despite the limited budget for law enforcement in the sparsely populated county, according to GovTech. Meanwhile, police in a California town are using AI tech “to transcribe video recordings taken on officers' body-worn cameras and create a first draft of a police report,” according to another GovTech report.
Research on why so many incompetent men rise to leadership positions found there’s a lot of antimeritocratic and implicit positive discrimination going on that favors not just men but overconfident, narcissistic, and incompetent men when it comes to leadership roles.
Often, even when women are appointed to very senior leadership roles, it isn’t because people have embraced what they bring to the table in terms of EQ, self-awareness, self-control, integrity, humility, people skills, et cetera. Rather, it’s because they go for a profile of somebody who may be biologically female but out-males males in masculinity. So there’s a queen bee or Margaret Thatcher phenomenon. In fact, there are many countries in the world that are run by women who look more alpha male than their male competitors. The point is not to have more biological women in charge but to have better leaders in charge.
Should You Write with Gen AI – Harvard Business Review
To Use AI or Not to Use AI? A Student’s Burden – Inside Higher Ed
The distinct human writer becomes more essential – Harvard’s Nieman Lab
How Indigenous engineers are using AI to preserve their culture – NBC News
Harvard Is Releasing a Massive Free AI Training Dataset of nearly 1 million public-domain books – Wired
How to identify AI-generated text: 7 ways to tell if content was made by a bot – Mashable
Over half of longer English-language posts on LinkedIn are AI-generated – Wired
TV Writers Found 139,000 of Their Scripts Trained AI – The Ankler
Is Grammarly AI? Notre Dame Says Yes – Inside Higher Ed
Stanford Professor Accused of Using AI to Write Expert Testimony Criticizing Deepfakes – Gizmodo
AI Companies Are Trying to Get MIT Press Books – 404Media
There’s No Longer Any Doubt That Hollywood Writing Is Powering AI – The Atlantic
What genre of writing is AI-generated poetry? - Statistical Modeling, Causal Inference, and Social Science
What Kind of Writer Is ChatGPT? – New Yorker
HarperCollins Confirms It Has a Deal to Sell Authors' Work to AI Company - 404Media
Why Watermarking Text Fails to Stop Misinformation and Plagiarism – Data Innovation
National Novel Writing Month faces backlash over allowing AI: What to know – Washington Post
How to Tell If What You're Reading Was Written By AI - Lifehacker
Can a Start-Up Help Authors Get Paid by A.I. Companies – New York Times
Google unveils invisible ‘watermark’ for AI-generated text – Nature
The Difference Between Editing Human vs AI Writing - Rebecca Dugas on Substack
Writer Ted Chiang on AI and grappling with big ideas – NPR
AI is My Research & Writing Partner. Should I disclose it? – Wired
Writers Guild Calls on Studios to Take “Immediate Legal Action” Against AI Companies – Hollywood Reporter
You can now ask Claude to mimic your writing style – Tech Radar
University of Reading researchers “slipped 63 AI-generated submissions into the school’s examination system. Even with no editing or efforts to hide the AI usage, 94 percent of those went undetected, and nearly 84 percent got better grades than a randomly selected group of students who took the same exam. The issue with such tools is that they usually perform well in a lab, but their performance drops significantly in the real world. ArsTechnica
At some point we all decline the opportunity to discover what really is the case. We willingly give up a shot at learning the truth about the world out of fear that it will expose truths about ourselves, especially our insufficient courage for self-examination. We prefer the illusion of self-reliance and embrace our ignorance for no other reason than it is ours. It doesn’t matter that reliance on false opinion is the worse sort of dependence. It doesn’t matter that through stubbornness we might pass up a chance at happiness. We prefer to go down with the ship rather than have our names scraped off its hull.
Mark Lilla, Ignorance and Bliss: On Wanting Not to Know
What: This webinar series will help you discover open and hidden sources, refine your investigative skills, and learn how to navigate information challenges in climate journalism.
Who: Iryna Ponedelnik Project Manager
When: 6 am, Eastern
Where: Zoom
Cost: Free
Sponsor: Network for Border Crossing Journalism
What: We'll explore the fascinating neuroscience behind chatbot learning, walk you through the step-by-step process of designing your own chatbot, and equip you with essential dos and don'ts for effective implementation.
Who: Margie Meacham is an expert in adapting AI technology to accelerate learning and support performance. She teaches training organizations around the world how to leverage AI for education and training.
When: 3 pm, Eastern
Where: Zoom
Cost: Free
Sponsor: Training Magazine Network
What: This webinar series will help you discover open and hidden sources, refine your investigative skills, and learn how to navigate information challenges in climate journalism.
Who: Iryna Ponedelnik Project Manager
When: 6 am, Eastern
Where: Zoom
Cost: Free
Sponsor: Network for Border Crossing Journalism
What: Share your questions about Bluesky, what you’re thinking about in this space, the successes/challenges of your experimentation on it.
Who: Bluesky’s Emily Liu
When: 4 pm, Eastern
Where: Zoom
Cost: Free
Sponsor: Online News Association
The highest thing that one man can do for another is simply to make him restless.
The GPT era is already ending something has shifted at OpenAI – The Atlantic
Harvard Is Releasing a Massive Free AI Training Dataset of nearly 1 million public-domain books – Wired
Their Job Is to Push Computers Toward AI Doom – Wall Street Journal
Google Unveils A.I. Agent That Can Use Websites on Its Own – New York Times
Former OpenAI researcher raises $40 million to build more empathetic audio AI – Reuters
Meta rolls out internal AI tool as it pushes into business market – Financial Times
Perplexity expands its publisher program – Tech Crunch
The UC Berkeley Project That Is the AI Industry’s Obsession – Wall Street Journal
The Furious Contest to Unseat Nvidia as King of A.I. Chips – New York Times
OpenAI looks at chatbot ads – Axios
Amazon Announces Supercomputer, New Server Powered by Homegrown AI Chips – Wall Street Journal
OpenAI explores advertising as it steps up revenue drive – Financial Times
Amazon eyes news partners for revamped AI Alexa voice assistant – Axios
Labelers training AI say they're overworked, underpaid & exploited by big US tech companies - CBS News
Who's winning the AI race – Axios
OpenAI hits pause on video model Sora after artists leak access in protest – Washington Post
The AI War Was Never Just About AI – The Atlantic
Stanford’s AI Center names US the top AI ecosystem, China follows – Semafor
The future of Windows is cloud and AI – The Verge
TV Writers Found 139,000 of Their Scripts Trained AI – The ankler
Five Canadian news media outlets sue OpenAI for copyright breach – Al Jazeera
Biden’s Farewell to China’s Tech Sector: A New Type of Forbidden Chip - Wall Street Journal
When failure becomes invisible, the difference between failure and success may also become invisible. -David McRaney
Creative people who can “think out of the box” are prized in the business world, the arts, and science. But a new study has found that creative thinkers are also more likely to cheat to get ahead, and to rationalize away less-than-ethical behavior. Harvard Business School researchers gave personality quizzes to hundreds of study participants and then asked them to perform quick games or other tasks for cash. Participants who scored high on a creativity test were more likely to falsify their results so they could earn more prize money. People who were merely high in intelligence, however, were not more dishonest. It appears that the same “divergent thinking” and “cognitive flexibility” that enable creative people to come up with innovative ways of looking at things also equip them to circumvent ethical norms—and to justify their cheating to themselves. “When you’re a creative person, you can use that creativity to come up with reasons for why unethical behaviors may be okay,” researcher Francesca Gino tells The Boston Globe. These “self-serving rationalizations,” she said, can include deciding that “other people would cheat under the same circumstances or that a little cheating will not hurt anyone.”
The Week magazine
Researchers Use AI To Turn Sound Recordings Into Accurate Street Images – University of Texas
Samsung has developed an audio eraser feature for smartphones that will allow users to erase unwanted sounds from videos – Data Company
Former OpenAI researcher raises $40 million to build more empathetic audio AI – Reuters
The Most Hyped Bot Since ChatGPT Remember Sora? – The Atlantic
OpenAI’s video generator, Sora, aims to kickstart the AI video era – Washington Post
NVIDIA's new AI model Fugatto can create audio from text prompts & modify existing sound files - Engadget
Randy Travis’s beautiful baritone was lost. AI helped him sing again. - Washington Post
Polish radio station ditches DJs, journalists for AI-generated college kids – The Register
Adobe Firefly Video Model: How AI is Changing the Future of Video Editing - Unite
There’s a New Hit Podcast That Will Blow Your Mind: The hosts aren’t human. – Wall Street Journal
Podcast: AI and Voice Replication - Illusion of More
Adobe’s AI video model is here, and it’s already inside Premiere Pro – The Verge
Talking through AI and the future of music with will.i.am – Semafor
Amazon is allowing Audible narrators to clone themselves with AI - The Verge
This Hit Music Radio Station Is Fully AI-Generated – Radio World
Amazon's AI Generator Tool Can Now Create Audio Ads – AdWeek
How To Create And Customize An AI Podcast With Google’s NotebookLM – Forbes
Zoom will let AI avatars talk to your team for you - The Verge
Mariah Carey Responds to Claims Her Spotify Wrapped Video Was Made With AI – Hollywood Reporter
Happiness doesn't depend on how many external blessings we have snatched from life. It depends only on our attitude toward them. -Aleksandr Solzhenitsyn (born Dec. 11, 1918)
Natural language processing - This is type of machine learning that makes human language intelligible to the machines. The first step is tokenization. Text is divided into units called tokens and then transformed into vectors—where the words are represented by numbers. These word vectors are lists of numbers. More than 1,000 numbers can be used to represent a single word—meaning that word vector has a high dimension—it’s very nuanced. A low dimension for a word vector means the list of numbers is low—not as nuanced but easier to work with. A deep learning model (typically a transformer model) uses these vectors to understand the meaning of words and determine how they relate to one other. For instance, king would relate to man while queen would relate to woman.
More AI definitions here
I dwell in possibility… –Emily Dickinson (born Dec. 10, 1830)
Our brains mix real imagery with mental and emotional baggage, which affects performance. Slugger Mickey Mantle is reported to have once said after hitting a long home run, "I just saw the ball as big as a grapefruit." In contrast, poor hitters may see the baseball as small. It’s not just out of reach for them physically but emotionally as well.
A Purdue University study tested the kicking ability of more than 20 athletes who don't play football. They were asked to estimate the size of the goalposts before and after each of 10 attempts to kick a field goal. The more successful the athlete, the more likely they were to overestimate the size of the posts and underestimate the distance.
Success biased the kickers’ perception of the difficulty of their task. Professor Jessica Witt says, “Before you kicked, you really didn’t know what your abilities were going to be.’’ She found the same effect in past experiments with softball players and golfers. University of Virginia psychologist Dennis Proffitt has put together tests that show the effect holds true even when it comes to dangerous situations.
Which are you imagining in your life—success or failure?
Stephen Goforth
For the foreseeable future, we’ll still need pro video editors who master the technical details of visual storytelling. But for many everyday situations — trimming a meeting recording, pulling social media clips, or gathering quick highlights — natural language editing may soon be a widely-adopted accelerator of the process. It’s not mature yet, but it’s poised to make video editing accessible to everyone who can describe what they want. AI is beginning to democratize creative work that used to require technical expertise.
Jeremy Caplan of WonderTools
Tokenization - The process where an LLM creates a digital representation (a token) of a real thing—everything gets a number; words are translated into numbers. Think of a token as the root of a word. “Creat” is the “root” of many words, for instance, including Create, Creative, Creator, Creating, and Creation. “Create” would be an example of a token. Examples
More AI definitions here
Becoming is a service of Goforth Solutions, LLC / Copyright ©2025 All Rights Reserved