In partnership with
The CBD Brand that Actually Does What It Says
CBD is having a moment and CBD brands are everywhere right now. But with so many options, it can be tough to know what to look for – and which brands actually follow through on their claims.
That’s where CBDistillery sets itself apart. For the past decade, all of CBDistillery's products have been made with non-GMO, US-grown industrial hemp and with 100% clean ingredients – no corn syrup or artificial colors, flavors, preservatives or sweeteners.
They frequently introduce new and innovative products formulated to meet a variety of wellness needs and customer preferences. And every single batch of CBDistillery’s gummies, oil tinctures, topicals and softgels are tested by an independent, accredited lab to make sure what’s in the package matches what’s on the label.
Right now, CBDistillery is offering our readers a special deal: save 50% on orders of $75 or more with code NEWS50!
Beginners in AI
Good morning and thank you for joining us again!
An AI system just settled a quantum computing question that stumped experts for 18 years, and a computer checked every step of its proof. Today's lead explains what was proven in plain words, and why it doesn't mean your passwords or bank are in trouble. As always, every story here is human picked and edited.
THE FRONT PAGE
An AI Cracked a Quantum Question Open Since 2008, and a Computer Checked It
TLDR: A research system running an unreleased Anthropic model proved the Aaronson-Ambainis conjecture, a quantum computing puzzle open since 2008, and a computer proof checker accepted every step.
The Story:
Physical Superintelligence (PSI), a small research startup, says its system wrote a complete proof on day three of a run, then improved it twice. The puzzle asks how much faster a quantum computer can be on random questions with no hidden pattern. The answer is faster, but not wildly faster. The huge speedups people talk about, like cracking codes, only show up when a problem has structure to exploit.
Lean, a program that checks math line by line, signed off on the proof. A Stanford team that solved a smaller piece in August has read it and called it "a huge result." On Sunday, Scott Aaronson, who first posed the question with Andris Ambainis, wrote on his blog that he's "disappointed that humans didn't get to snag this conjecture before the robots did." It hasn't been through formal peer review yet.
Its Significance:
This is the second famous computer science question an AI system has knocked down in two weeks, after OpenAI said in late October that its AI showed the Exponential Time Hypothesis doesn't hold. Aaronson says open problems are now "dropping like flies." Don't read it as "quantum computers are dead," though. He points out it rules out one narrow kind of speedup that experts already doubted, so code-breaking and chemistry simulations are untouched. And a person still steered it: PSI's David Turturean ran the system and is responsible for the paper being right.
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. There's even a cinematography course coming out soon.
QUICK TAKES
The story: Bloomberg's Mark Gurman reports Apple is exploring a Siri that chases a goal and keeps working without step-by-step prompts, the way OpenAI's Dots, Meta's Muse and xAI's Grok do. He says it could come as early as 2027, after CEO John Ternus told software chief Craig Federighi to move faster than basic Siri fixes.
Your takeaway: It's a report, not an announcement, so nothing changes on your iPhone today. Siri can already do tasks inside Apple's own apps, and an agent version would mean handing it a whole errand instead of one command at a time.
The story: Seattle's Firmly raised $15 million, led by Squarepoint Capital with J.P. Morgan Payments, Mastercard and Samsung Next joining, so stores can sell inside AI shopping assistants without writing code. Its Ask Agent shopping chat, which a store adds with one line of code, raised sales by up to 18% in early tests, the company says.
Your takeaway: Big payment companies are paying for the pipes that let an AI buy things for you. That's handy, but Visa's new test in Trending shows why you should still check the cart before it pays.
The story: Employers tied 1,265 October layoffs to AI, under 3% of the 43,994 total and the lowest monthly count this year, says outplacement firm Challenger, Gray & Christmas. AI is still behind 121,401 cuts so far in 2026, about 20%, but "market and economic conditions" has now passed it as the top reason.
Your takeaway: One quiet month isn't a trend, so learning the basics is still cheap insurance. Our guide to free AI courses for beginners lists 12, and the shortest takes about 70 minutes.
TOOLS ON OUR RADAR
📈 Desmos Free: Graphing, scientific, geometry and 3D calculators in your browser, handy if today's lead made you curious about math.
🗺️ draw.io Free and Open Source: Builds flowcharts and diagrams with no account, a free swap for Visio or Lucidchart.
✂️ LosslessCut Free and Open Source: Trims the boring parts out of long videos in seconds, because it never re-encodes them.
🖨️ NAPS2 Free and Open Source: Scans paper into searchable PDFs, reading text in over 100 languages, with no ads.
✉️ Tuta Mail Freemium: Encrypted email and calendar, a private swap for Gmail with 1 GB free.
🎞️ VEED Freemium: Edits videos in your browser and writes subtitles automatically, with a free plan to start.
TRENDING
Nvidia's Board Investigates Reports Its CEO Was Targeted by a Foreign Spy Effort. Independent directors and outside lawyers will review reporting that Jensen Huang was the target of a foreign intelligence effort. The board says it has no reason to believe national security or Nvidia's secrets were compromised, and Huang says he'll cooperate.
Visa Tested 22 AI Shopping Agents, and They Failed 42.8% of the Time. In more than half the failed runs, the AI bought something anyway, either the wrong product or before learning what the shopper needed. Per Fortune, Google and Meta agents tied for the top score at 82%, a Claude agent came third at 76% and an OpenAI agent sixth at 64%.
OpenAI Teams Up With the Hackers Who Broke Into Its Systems. This summer, security firm Hacktron got from OpenAI's community forum to its private code in under 72 hours, with Claude as a helper, and collected a $6,500 bug bounty. Now the two will pick test targets together, starting with how AI tools connect to outside services.
Hundreds of AI Hacking Tools Were Left Wide Open on the Web. Censys found more than 800 copies of Artex, an AI agent that can run an attack with little human help, reachable without a password. Many attacks linked to it got in through weak passwords and no two-step login, so turning on 2FA still blocks a lot.
A Chinese AI Video Startup Raised $200 Million and Hit a $1 Billion Value. Beijing's Sand.ai makes the Magi video models, and its VidMuse app passed $10 million in yearly revenue within two months of launching in January.
AI Industry Money Mostly Lost in the Midterms, on Both Sides. AI super PACs spent $79 million across 68 races. Leading the Future, backed by people at Andreessen Horowitz and OpenAI's Greg Brockman, lost its biggest bet in South Carolina, and both Republicans that Anthropic-backed Public First supported lost too.
TRY THIS PROMPT (copy and paste into Claude, Grok, ChatGPT, or Gemini)
🔍 Proof Check: Paste an "AI solved it" headline and see how far it's been checked
Build a single-file HTML app called Proof Check in vanilla HTML, CSS and JavaScript. No API key needed.
I paste a headline claiming an AI solved a math, science, medical or security problem, or tap one of four example buttons.
The app calls the Anthropic Messages API with the web_search tool to find the original announcement, any paper, any computer proof check, named outside experts who reviewed it, and whether it's been peer reviewed. It must never invent names, dates, quotes or links, and it says so plainly when it can't confirm something.
Show a status badge (claimed, partly checked, well checked, disputed), a five-step bar from "claimed" to "peer reviewed," a plain-English summary, a "who checked what" list, what's still open, and real source links.
Style: deep navy background, violet and gold accents, Unbounded, Lora and JetBrains Mono fonts.What this does: Paste any big "AI solved it" headline, like today's quantum proof. The app searches for the paper, any computer proof check, outside experts and peer review, then shows how far the claim has been checked so far. You get a status badge, a five-step bar from "claimed" to "peer reviewed," and a list of what's still open, with real source links.
The holiday streaming season is almost here
Are you ready? As revealed in our holiday playbook, Roku Ads Manager lets you easily set up, measure and optimize your self-serve holiday ad campaign. You can even create a shoppable ad unit that lets streamers buy directly from their remote.
WHERE WE STAND(based on today's news)
✅ AI Can Now: Write a full proof for a math problem experts left open for 18 years, then pass a computer check of every step.
❌ Still Can't: Do that research on its own. PSI's David Turturean ran the system, and he answers for the paper being right.
✅ AI Can Now: Shop for you and ask questions to learn what you need, with the best agents passing 82% of Visa's tests.
❌ Still Can't: Reliably stop when nothing fits. In more than half of Visa's failed runs, the agent bought something anyway.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
Hear the man behind today's conjecture explain what a quantum computer is and isn't, in a 2020 talk that runs about an hour and a half. He spends a good stretch on overhyped quantum claims, the same caution he's urging now that an AI has settled his puzzle.
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.
