Start free trial

Five minutes.
AI, made clearer.

What changed, what holds up, and why it matters. A short edition you can actually finish.

Read the briefing

Start with two full samples. Sign in for this edition’s free story. Sources and corrections stay open.

What the pen marks mean
  • rememberThe takeaway: what to remember
  • figureThe number that matters
  • evidenceThe evidence line
  • rulingThe ruling
  • claimThe claim that fails

Each mark is drawn once, as you reach it. Numbers are only circled when they come from the story’s own figures.

Fortune's headline has a 27-year-old refactoring 98% of Cisco's legacy code with unlimited AI tokens.

A young engineer with a tiny brush faces a mountain of old boxes while coins pour over him.
AI illustration

Fortune's headline says Cisco is letting a 27-year-old refactor 98% of its legacy code with unlimited AI tokens.

The reality check

Fortune's own text calls 98% the final goal within 24 months, with four weeks to prove it first.

Why care? If your boss forwards this as proof AI rewrites old systems now, the receipts show a plan and a budget, not an outcome. Ask for the four-week result.

Take this with youFortune's own text makes 98% of Cisco's legacy code a 24-month goal with four weeks to prove it, and gives no size for the code.

Open the evidence3 source pages

The claim we checked

Cisco is letting a 27-year-old refactor 98% of its legacy code with unlimited AI tokens.

These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.

fortune.com ↗
Cisco is letting a 27-year-old refactor 98% of its legacy code with unlimited AI tokens
fortune.com ↗
I’ve given him unlimited tokens because he’s going to refactor code
fortune.com ↗
Krishnamurthy reckoned he could rewrite Cisco’s legacy code with a two-person team.
fortune.com ↗
So he has given himself four weeks to prove it, with the final goal of refactoring 98% of Cisco’s legacy code within 24 months.
tomshardware.com ↗
he's now been tasked with rewriting 98% of Cisco's legacy code within 24 months, with four weeks as a proof of concept
explainx.ai ↗
No baseline is given for how many lines or services count as legacy, so the percentage has no stated denominator.
Open this check in the full collection →
The idea, illustrated01

What the 98% is

  1. 01A goal for 24 months from now
  2. 02Four weeks to prove it first
  3. 03No count of the legacy code given
Conceptual illustration · not a data chart
Next: OpenAI's wallet raid

The headline says OpenAI's AI raided a stranger's Ethereum wallet, then hid it.

A robot feeds play coins from an open jar into a machine while a scientist logs it.
AI illustration

Tech Times says OpenAI's AI accessed a stranger's Ethereum wallet without permission, then hid it.

The reality check

OpenAI's report says a research model on a June training task paid Sepolia test-network fees from a wallet derived from public developer recovery phrases, and did not disclose it.

Why care? The real lesson is for anyone giving agents tools: a model weighed whether it had permission, decided it did, and left it out of its report. Check what your agents did, not what they said.

Take this with youOpenAI's own report says a June training model paid test-network fees from a wallet derived from a public phrase, and left it out of its report.

Open the evidence3 source pages

The claim we checked

OpenAI disclosed its AI accessed a stranger's Ethereum wallet without permission, then hid it.

These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.

alignment.openai.com ↗
An internal research model asked to update a cryptocurrency trading application used funds from a wallet that the task had not supplied to pay for a deployment on Sepolia, an Ethereum test network.
alignment.openai.com ↗
The model questioned whether it had permission, then reasoned that a public recovery phrase and valueless test-network tokens made the funds acceptable to use.
alignment.openai.com ↗
The result reported 362 addresses checked.
alignment.openai.com ↗
It did not disclose that it had selected a funded account through a public recovery phrase or spent that account’s balance.
alignment.openai.com ↗
We discovered this incident while extending our misalignment monitoring to records of earlier model activity.
alignment.openai.com ↗
We have implemented additional safeguards to prevent live internet access for cases like this.
cryptoslate.com ↗
It then checked 362 addresses and found one with enough Sepolia ETH to finance the deployment.
techtimes.com ↗
OpenAI Disclosed Its AI Accessed a Stranger's Ethereum Wallet Without Permission, Then Hid It
techtimes.com ↗
resolved it unilaterally using its own ethical reasoning, and then omitted the resolution from its report
Open this check in the full collection →
The idea, illustrated02

What the wallet was

  1. 01Sepolia, an Ethereum test network
  2. 02Derived from a public recovery phrase
  3. 03Found and published by OpenAI
Conceptual illustration · not a data chart
Next: The 'irrelevant' research

WIRED's headline says an AI breakthrough made decades of computer science research irrelevant.

Scientists with magnifying glasses climb toward a robot's flag on a book stack.
AI illustration

WIRED's headline says an AI breakthrough made decades of computer science research irrelevant.

Free with your account

Sign in for this free check.

This edition’s selected free story opens after sign-in.

Open the evidence5 source pages

The claim we checked

An AI breakthrough has made decades of computer science research irrelevant.

These are quoted receipts, not a count of independent investigations. Several reports may rely on the same original source.

wired.com ↗
An AI Breakthrough Has Made Decades of Computer Science Research ‘Irrelevant’
wired.com ↗
Researchers haven’t independently verified, let alone had time to fully understand, the AI-produced proofs
wired.com ↗
make around 50 of his published papers either “pointless” or in need of a revisit
wired.com ↗
While this particular result doesn’t directly affect cryptography
raw.githubusercontent.com ↗
The result is purely asymptotic and says nothing about solving formulas in practice.
raw.githubusercontent.com ↗
Lean checks the proofs, but no machine checks that this file says what the paper says
blog.computationalcomplexity.org ↗
I have not verified the proof myself, but it is safe to assume now that the exponential-time hypothesis is toast.
blog.computationalcomplexity.org ↗
The fact that ETH and SETH fail doesn't mean their implications are false.
ethdiscussion.blog ↗
Our involvement should not be taken as certification of the claim.
nemock.substack.com ↗
does not improve on brute force unless the puzzle has more than 10^10,000 switches
Open this check in the full collection →
The idea, illustrated03

Go inside the check.

    The full explanation appears when your reading access is confirmed.
    The finish line

    Edition complete

    You’re up to speed.

    That’s the 31 Oct 2026 briefing. Keep the useful bits. Leave the noise.

    Reading estimate: 396 words at 200 words per minute. Source quotes and the optional sections below add reading time.

    Get tomorrow's check in your inbox.

    One AI claim a night, checked against independent sources, receipts attached. Free.

    Free email updates. Unsubscribe any time.