Bloodbath, Bounce, and Builders: Thursday's selloff, Friday's jobs surprise, and the AI report that changes the clock

Thursday was one of the worst days the market has seen since April 2025. Friday gave some of it back. And in the middle of all that noise, a quiet report landed that matters more than either day.

Thursday took trillions

Thursday, November 5, the market fell apart in a few hours. The Dow Jones Industrial Average dropped 1,833 points, or 3.4 percent, and closed at 51,523. The S and P 500 fell 4.4 percent to 7,353. The Nasdaq Composite fell 5.5 percent to 24,315, its worst day in 19 months, and more than 10 percent below its recent record high. That puts the Nasdaq in correction territory.

Every one of the 11 sectors in the S and P 500 finished the day in the red. Chip stocks led the fall. Nvidia dropped 9.19 percent. AMD dropped 17.43 percent. The trigger was a simple question investors finally asked out loud: the big cloud companies are spending enormous sums on AI, and when does that spending start paying for itself? Global markets lost about 5.5 trillion dollars in value across November 5 and 6. Nothing about the technology changed on Thursday. What changed was the story investors told themselves about how fast the money comes back.

Friday brought a jobs surprise

Friday morning at 8:30, the Bureau of Labor Statistics released the October jobs report, and it was strong. Employers added 165,000 jobs in October. Forecasters expected 59,000 in the Dow Jones consensus and 65,000 in the Reuters poll. The unemployment rate fell to 4.1 percent from 4.2 percent. Average hourly earnings rose 0.4 percent for the month to 38 dollars, up 3.1 percent from a year ago.

The prior 2 months were revised up by a combined 84,000 jobs. August now stands at 186,000, up from 133,000. September now stands at 60,000, up from 29,000. Professional and business services led October with 70,000 new jobs. Health care added 42,000.

Stocks bounced on the news, but only partway. Late Friday, the Nasdaq was up 1.45 percent to 24,667 and the S and P 500 was up 0.62 percent to 7,399, while the Dow slipped 0.27 percent to 51,382. Small caps kept falling. The Russell 2000 dropped 4.45 percent to 2,729. So Friday was a recovery for big tech and a warning for everything smaller. The jobs report says the economy is hiring. The market says it still does not trust the AI spending bill.

Bitcoin held its ground, barely

Crypto did not panic, and it did not rally either. Bitcoin sits at 75,789 dollars, down 0.36 percent in the last 24 hours. Ethereum is at 2,227 dollars and 58 cents, down 0.76 percent. The total crypto market is worth about 2.57 trillion dollars. Bitcoin holds 59.24 percent of that total. The fear and greed index reads 41, which is fear.

After a week when stocks lost trillions, Bitcoin barely moved. It did not rescue anyone either. My view has not changed. Bitcoin is a long term position for people who understand what they own, not a trade to make because the news looks scary. Do not buy it on a red day because you feel brave, and do not sell it on a red day because you feel afraid.

AI is now improving AI, about twice as fast

This is the story that matters most this week, and it got the least attention. On November 5, METR, an independent nonprofit that evaluates AI systems, published an assessment of Anthropic, the company behind Claude. METR spent 11 weeks inside Anthropic, from August 12 to October 29, 2026, looking at internal data, interviewing staff, and measuring one question: is AI speeding up the work of building better AI?

The answer is yes, by a lot. METR estimates that Anthropic is now improving its models about 1.9 times as fast as it was before October 2025. The 95 percent confidence range is 1.5 to 2.8 times. In plain terms, Anthropic packed roughly 2 years of old progress into the last year. METR estimates that about 80 percent of that speedup comes from AI itself, not from bigger computers.

Let me explain what they actually measured. METR used an index called AECI,

Anthropic’s version of the Epoch Capabilities Index. It turns benchmark test results across math, software engineering, science, multimodal tasks, and language into one score. Before October 2025, that score rose about 15 points a year. After October 2025, it rose about 29 points a year. That ratio is the 1.9 times.

Now the honest caveats. Anthropic redacted a majority of the evidence METR found most informative, so the public cannot check the most sensitive parts. METR says it would not be surprised if the true figure were closer to 1.3 or 3 times the old rate. And the viral version of this story is wrong. METR is explicit that AI is not yet able to make highly significant and fully autonomous contributions to the most important parts of model training.

What does this mean for you? The tools you are learning today will be meaningfully better in 6 months. The gap between a beginner and a builder is shrinking because the tool does more of the heavy lifting, but only if you actually use it.

Here is a workflow you can copy today, starting from zero. Pick one small thing you want to build, like a simple website or a spreadsheet that tracks your spending. Open Claude and describe what you want in plain English. Ask it to build the first version, then ask it to explain every part it made, line by line, until you understand it. Then change one thing yourself and ask it to check your change. That loop, build, explain, change, check, is how you learn. It costs nothing but time.

One more step. Verify what the AI gives you. Run the code. Check the number against a real source. The people who will win in the next 2 years are not the people who trust AI the most. They are the people who use it the fastest and check it the hardest.

What to do if you are starting from zero

Thursday and Friday gave you the whole lesson in 2 days. Markets fall hard on a story. Jobs data can surprise everyone. Bitcoin can sit still while the world shakes. And AI is quietly getting better at building itself, twice as fast as last year. If you are starting from zero, do not try to time any of that. Build the skill instead. Learn one tool deeply. Ship one small thing. Then ship another.

The market will keep having Thursdays. Your skill is the only asset that does not drop 5.5 percent in a day.

Here is the uncomfortable truth. Most people reading this will nod, close the tab, and change nothing. They will wait for a better time, a clearer signal, a calmer market. That time never comes. Thursday proved it. The people who were ready on Thursday were the people who built something before Thursday. So start today. Open the tool. Build the small thing. Break it. Fix it. That is the whole secret, and it has never changed.

Join my free Skool community: https://www.skool.com/henry-jenkins-5224
My website: https://henryljenkins.com
LightofLife.global
Twitter @Nevaquit
YouTube @HenryJenkins3

Henry Jenkins