In partnership with
Speak fuller prompts. Get better answers.
Stop losing nuance when you type prompts. Wispr Flow captures your spoken reasoning, removes filler, and formats it into a clear prompt that keeps examples, constraints, and tone intact. Drop that prompt into your AI tool and get fewer follow-up prompts and cleaner results. Works across your apps on Mac, Windows, and iPhone. Try Wispr Flow for AI to upgrade your inputs and save time.
Beginners in AI
Good morning and thank you for joining us again!
Welcome to this daily edition of Beginners in AI, where we explore the latest trends, tools, and news in the world of AI and the tech that surrounds it. Like all editions, this is human curated and edited, and published with the intention of making AI news and technology more accessible to everyone.
THE FRONT PAGE
AI Is Now Rewriting What 85 Million Roblox Players Say, but Kids Are No Safer
TLDR: Roblox is rolling out AI that rewrites players' curse words into cleaner language in real time, but critics say the company is focused on policing swear words while still failing to protect kids from adult predators.
The Story:
If you've ever played Roblox, you've seen the "####" that pops up when someone types something the filter catches. Now Roblox is using AI to rephrase those messages instead of just hiding them. So "Hurry TF up!" becomes "Hurry up!" and everyone in the chat gets a note saying the message was cleaned up. The feature works across all supported languages for age-verified users. But here's the problem: Roblox is facing lawsuits from at least seven states accusing the platform of being a "playground for predators." At least 30 people have been arrested since 2018 for abducting or sexually abusing children they first contacted on Roblox. Nebraska's attorney general filed suit just one day before this AI chat feature was announced.
Its Significance:
Roblox has about 85 million daily players, and roughly 40% of them are under 13. Parents have been begging for better protections against predators who use the platform to groom children, often luring them to other apps like Discord. Instead, Roblox's big safety announcement this week is an AI that turns curse words into nicer phrases. It doesn't address the core concern that adults can still interact with children on the platform in ways that have led to real-world abuse. When your platform is being compared to "a creepy cargo van at the edge of a playground" by a state attorney general, using AI to add a profanity monitor seems like more of a distraction than anything.
QUICK TAKES
The story: Researchers from IonQ and Microsoft have proposed using quantum computers to generate ultra-accurate data about how electrons behave in molecules, then feeding that data to AI models running on regular computers. The trained AI could then make fast predictions about new materials and drugs without needing the quantum machine every time.
Your takeaway: This is a clever shortcut. Quantum computers aren't powerful enough yet to do all the heavy lifting on their own, but they can teach AI to be way more accurate. It could speed up how we discover new medicines and materials.
The story: February's jobs report showed 92,000 jobs lost and unemployment jumping to 4.4%. RedBalloon CEO Andrew Crapuchettes says AI is a big reason why: companies are using AI to boost productivity so much they don't need as many workers, and AI hiring tools are filtering out qualified applicants before a human ever sees their resume.
Your takeaway: Crapuchettes says the number one skill employers are asking for right now, across every industry, is being "AI-enabled." If you're not learning AI tools for your job, you could be invisible to hiring systems without even knowing it.
The story: Grammarly's "Expert Review" feature offers AI writing advice "inspired by" real people, including The Verge's editors, New York Times writers, and even deceased professors. None of them gave permission to be included. The feature also frequently crashed and sometimes linked suggestions to the wrong person's work entirely.
Your takeaway: This is a growing problem with AI tools: companies are using real people's names, reputations, and work to sell products without asking first. If it feels like identity theft with a polished UI, that's because it basically is.
TOOLS ON OUR RADAR
🐧 FreeTube Free and Open Source: A privacy focused desktop video player that allows you to watch your favorite online creators and subscribe to their channels without being tracked by corporate advertising algorithms. (Alternative to YouTube)
🍳 ChefGPT Freemium: A smart culinary assistant that asks what ingredients you currently have in your kitchen and instantly generates delicious step by step recipes to prevent food waste.
📞 * Answrr Freemium: An artificial intelligence virtual receptionist that answers your business calls around the clock booking appointments and handling customer inquiries with an incredibly realistic human voice.
📸 Popsa Freemium: An artificial intelligence photo printing application that automatically curates your best smartphone pictures and designs beautiful physical photo books in just a few minutes.
TRENDING
Microsoft Says Anthropic's Claude Can Stay in Its Products Despite Pentagon Blacklist - Microsoft's lawyers concluded that Anthropic's AI models can remain available to all customers except the Department of War, after the Pentagon designated Anthropic a supply chain risk. Google and Amazon followed with similar statements. Anthropic plans to challenge the designation in court.
How an Intern Helped Build AlphaGo, the AI That Beat the World's Best Go Player - Chris Maddison was a master's student when he helped kick off Google DeepMind's AlphaGo project. He says the core technology behind AlphaGo (predicting the next move, then refining with reinforcement learning) is the same approach that powers today's chatbots like ChatGPT.
Breacher.ai Adds Live Deepfake Video Calls to Its Security Testing Platform - Companies can now test whether their employees would fall for a deepfake video call on Zoom, Teams, or Google Meet. The platform already offered fake phishing emails and cloned voice calls, and one security tester said it felt like "an episode of Black Mirror."
Fed Officials Are Split on How Fast AI Will Reshape Jobs and Inflation - Federal Reserve officials agree AI will cause major economic shifts, but they can't agree on how quickly. Some warn that displaced workers will face longer job searches and higher unemployment, while others think the long-term effects will be positive.
Seth MacFarlane Used AI to Transform Into Bill Clinton for 'Ted' Season 2 - After prosthetics and traditional CGI "looked terrifying," MacFarlane used AI to become a 1990s-era Clinton in a live-action comedy scene. The clip went viral, with viewers split between being impressed and concerned about AI replacing actors.
North Korea Is Using AI to Sneak Fake IT Workers Into Western Companies - Microsoft warns that North Korean groups are using AI-generated fake IDs, voice-changing software, and deepfake tools to get hired at Western companies under stolen identities. The paychecks go straight to funding the regime's military programs.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
Build a freelance rate calculator that turns your income goal into a confident hourly number with overhead and buffer built in
Build a freelance rate calculator app. Features: sliders for annual income target (20K-300K), billable hours/week (5-60), weeks off/year (0-12), overhead % (0-60), safety buffer % (0-50). Display: minimum hourly rate (no buffer), recommended hourly rate, day rate (8hrs), weekly rate, and cost breakdown (base income + overhead amount + buffer amount = total needed/year). Dark navy theme. 900px wide.What this does:
Drag the sliders to set your income target, weekly hours, time off, overhead, and safety buffer. The app instantly shows your floor rate, recommended rate with everything built in, plus day and weekly rates for project or retainer work.
What this looks like:
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Automatically rephrase inappropriate chat messages into cleaner versions while keeping the speaker's intent, in real time across multiple languages.
❌ Still Can't: Tell the difference between a qualified job applicant and an unqualified one well enough to replace human hiring judgment, leaving good candidates stuck in algorithmic limbo.
✅ AI Can Now: Transform a person's face into a convincing deepfake of someone else on a live video call, good enough that trained security professionals can't tell the difference.
❌ Still Can't: Generate accurate, reliable "expert" writing feedback. Grammarly's feature linked to wrong sources, crashed frequently, and attributed advice to people who never gave it.
FROM THE WEB
Compare the consistency in this video to AI videos from last year. Even the cost is going down.
Launching Tuesday, March 10th
RECOMMENDED LISTENING/READING/WATCHING
Former OpenAI researcher Leopold Aschenbrenner published this sprawling analysis of where AI is headed and why the next few years could be the most consequential in human history. He makes a detailed, data-heavy argument that superintelligence could arrive sooner than most people think. It's long, dense, and impossible to dismiss.
Great Docs Drive Real Revenue
Your documentation is the first thing developers evaluate before adopting your product. Mintlify helps you ship docs that accelerate adoption, reduce support load, and convert users into customers.
Thank you for reading. We’re all beginners in something. With that in mind, your questions and feedback are always welcome and I read every single email!
-James
By the way, this is the link if you liked the content and want to share with a friend.
Some * designated product links may be affiliate or referral links. As an Amazon Associate, I earn from qualifying purchases. This helps support the newsletter at no extra cost to you and Amazon makes a tiny hair less.
