Your K-Cup Has Been Sitting in a Warehouse for Months
That grocery store pod? It was roasted long before it hit the shelf. Coffee starts losing flavor within weeks of roasting — which is why your morning cup tastes flat, no matter how nice your machine is.
Angelino's does it differently. A third-generation roaster in Los Angeles, they roast, grind, and seal every pod in-house for peak freshness, then ship it to your door within days of roasting. The difference is the first thing you'll notice: real aroma, real flavor, no bitterness.
Mix and match from 50+ specialty coffees, teas, and flavored blends — all Keurig®-compatible — and save more with every box you add, up to 34% off at 12+. Order once whenever you're running low, or subscribe for an extra 5% if that's easier. Entirely up to you.
New customers get 15% off their first order — applied automatically, no code needed.
Beginners in AI
Good morning and thank you for joining us again!
A scientist's day is mostly reading papers and writing them up, and that's where AI helps the most right now. Today's lead is about OpenAI handing free ChatGPT to up to 200,000 researchers, what its first users say it saved them, and why that number deserves a careful read. As always, every story here is human picked and edited.
THE FRONT PAGE
OpenAI Will Give Free ChatGPT to Up to 200,000 Scientists Worldwide

TLDR: OpenAI is growing its free ChatGPT program for academic researchers toward 200,000 scientists, and a survey of its first users says their work would have taken nearly twice as long without AI.
The Story:
OpenAI said Monday it will hand out 50,000 more free grants to academic researchers, TNW reports. Each grant covers a researcher plus up to two collaborators, so up to 150,000 more people could get in. New grants come with ChatGPT Plus, not the pricier Pro plan pilot users got, though OpenAI says most pilot users stayed inside Plus limits anyway. Applications open November 23, first come, first served.
The pilot started in July and has 20,000 researchers in 33 countries. OpenAI also published a study of 1,819 of them and more than 580,000 of their messages: reading and learning made up 28% of their science chats, and writing 27%. Asked how long the same work would have taken without AI, 251 of them put it at about 1.9 times as long.
Its Significance:
Most of science is reading and writing, and those are the jobs today's chatbots handle best. A free seat for a grad student or a small lab could give back real hours every week. Read that 1.9x carefully, though. It's people's own guesses, with no stopwatch and no group working without AI to compare against, and OpenAI says itself the study doesn't prove AI speeds up discovery. You don't need a grant to try the reading part: paste a long report into any free chatbot and ask for a five-line summary plus the questions it leaves open.
JOIN THE COMMUNITY
Learn AI With Us
Most people try AI once, get a so-so answer, and quit. Our 14-day challenge gets you past that point, one short lesson a day.
Week one, AI starts working for you. It drafts a real email for you, flags which messages in your inbox need a reply, and turns messy notes into next steps. You also learn to catch its mistakes. By day 7 you have a weekly AI routine you'll want to keep.
Week two, you build something of your own. You find a skill people will pay for and turn it into an offer, with a one-page pitch, a short video and your first client emails, all drafted with AI and ready to test. Day 14 hands you a 30-day plan.
10 to 15 minutes a day. No tech background needed. We just released one for older adults, with a gentler progression in the lessons.
QUICK TAKES
The story: Type "make me a site" in Gemini Notebook's chat, describe what you want, and Google says it builds a clickable page from your sources, like an interactive periodic table, a history timeline or a project tracker. It's live now on the web for users 18 and up, with the phone app coming later.
Your takeaway: It's a quick way to turn study notes into something you can click through or share with a classmate. New to the tool? Our guide to NotebookLM, which Google renamed Gemini Notebook in July, shows how it answers only from the files you give it.
The story: Apple's new study app, built with medical AI company OpenEvidence, lets U.S. adults over 21 with an iPhone 15 or later ask health questions and get plain answers drawn from journals like JAMA, TechCrunch reports. Apple hopes to sign up as many as 200,000 people for more than a year.
Your takeaway: Apple says the data is anonymized and won't train AI models, and you choose whether to share things like sleep or lab results. It's a study, not a doctor, so use it to prepare questions for your next appointment.
The story: Jordan Belfort's startup Tone makes AI voices that shift their tone to persuade, like whispering to hint that something's running out, and he says they learned from 30 years of his sales material. It has raised about $10 million and brings in roughly $700,000 a year, though TechCrunch couldn't confirm the tech works as described.
Your takeaway: Belfort, who pleaded guilty to securities fraud in 1999, says Tone only calls people who opted in. Still, AI voices built to persuade are headed for your phone, so the old rule holds: never hand over card details on a call you didn't start.
TOOLS ON OUR RADAR
📚 Zotero Free and Open Source: Collects, organizes and cites your research sources for free, a handy sidekick to today's lead.
🔐 KeePassXC Free and Open Source: Keeps your passwords in an encrypted file on your own computer, a free swap for 1Password.
🔭 Stellarium Free and Open Source: Shows a realistic night sky for any place and time, so you can name what you see.
🗺️ Organic Maps Free and Open Source: Offline maps for hiking, biking and driving with no ads or tracking, a Google Maps alternative.
📊 LibreOffice Free and Open Source: A full office suite with spreadsheets, documents and slides, a free swap for Microsoft Office.
🌊 Flowith Freemium: One workspace for AI images, video, slides and websites, with 300 one-time free credits to start.
📰 beehiiv Freemium: Start your own email newsletter free for up to 2,500 subscribers, the platform this one runs on.
TRENDING
Elon Musk Joined 160,000 People Asking Nvidia to Bring Back Jensen Huang. Huang quit last Thursday after reports about his ties to a former customer executive U.S. officials suspect of spying for China, and the "Stand with Jensen" petition wants him back while an outside investigation runs. Treasury Secretary Scott Bessent says Nvidia should look outside for its next CEO, and the company reports earnings today, its first under interim CEO Colette Kress.
Nvidia's New Vera Rubin Chips Topped Every Test in a Major AI Training Benchmark. Nvidia says the system trained up to twice as fast per chip as the last generation and cut one big training run from 16 minutes to about four. It was the only platform with results on every MLPerf test this round, and some of its numbers are still waiting on outside checks.
Google's Gemini 4 Carbon Took First Place on a Hard Math Leaderboard, at 15 Times the Cost. It scored 85.4% on MathArena, ahead of Anthropic's Claude Fable 5.2 at 82.8% and OpenAI's GPT-6.1 Sol at 80.6%, The Decoder reports. GPT-6.1 Sol cost about $1 per run against $15.70 for Gemini, whose most common slip was running past the time limit.
A Startup That Says Its AI Handles 99% of Security Tasks Raised $125 Million. Corma, three months out of stealth, is now valued at $1 billion after NEA led the round with Sequoia and Coatue joining. The 99% figure is the company's own, and it says a shortage of computing power, not model quality, is what keeps its waitlist waiting.
Two Stanford Professors Launched Noeri With $110 Million to Build AI That Improves Itself. Yejin Choi and Carlos Guestrin's company tunes the tools and instructions around a business's AI, and a person has to approve every change before it goes live. Nvidia and Snowflake invested, but there are no named customers or prices yet.
TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)
🔬 Study Sniff Test: Check how much weight a "study shows" headline can hold
Build a single-file HTML app called Study Sniff Test in vanilla HTML, CSS and JavaScript. No API key needed. I paste a headline or claim that cites a study (like "scientists say ChatGPT nearly doubles their speed") and, if I have it, a link to the article.
The app calls the Anthropic Messages API with the web_search tool to find the original study, not just news coverage. It must never invent numbers, names or findings. Anything it can't find is marked "unknown," and if there's no original study it says so.
Show a verdict (Strong, Decent, Thin or Can't tell) with a strength bar, then cards for who ran it, what it measured, sample size, whether results were measured or self-reported, control group, peer review and who paid. End with whether the headline matches the study, questions worth asking, and source links.
Style: near-black background, mint and amber accents, Space Grotesk, Newsreader and JetBrains Mono fonts.What this does: Today's lead turns on one number scientists reported about themselves. This app checks any "a study shows" headline the same way: it finds the original study and tells you what was measured, how many people took part, whether anyone was compared against a group without the thing being tested, and whether the headline matches what the study found.
Stop Making Shoppers Spell It Out
“Running shos” shouldn’t return zero results. Neither should “red waterprof trail runners under $150.”
Traditional search needs exact keywords. Gladly understands what shoppers mean, typos and all, and turns dead-end searches into products that fit what they asked for.
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Help scientists with the reading and writing that fills most of their day. In OpenAI's pilot, reading made up 28% of science chats and writing 27%.
❌ Still Can't: Show it speeds up discovery. The 1.9x figure is scientists' own estimate, with no timer and no comparison group.
✅ AI Can Now: Score 85.4% on MathArena's competition math problems, Gemini 4 Carbon's new top mark.
❌ Still Can't: Keep an eye on the clock. Gemini 4 Carbon's most common failure was running past the time limit, once leaving a script running for more than 11 hours without checking on it.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
Wharton professor Ethan Mollick's 2024 book is a plain-spoken guide to working alongside chatbots, from writing to teaching to research. Mollick co-wrote the Boston Consulting Group study where consultants using GPT-4 finished 12.2% more tasks and worked 25.1% faster, one of the few tests that measured people against a group working without AI. It's a good next read if today's 1.9x number left you wondering what careful testing looks like.
How was today's edition? Tap one below, then tell me what you'd change on the next screen. I read every answer.
How was today's edition?
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.





