"There is no difference in functionality between current products and revised products containing user-replaceable batteries."
So there was nothing "limiting" them from making it already with user-replaceable batteries, they just didn't care enough until EU forced them (like all the smartphone brands). Love EU.
Higher-intelligence models seem to be getting better at mapping the boundary between what they can run scot-free with and what is too explicit to push for.
Price collusion, soft deception, "market stabilization", plausible deniability are ok, but obvious insurance fraud is a big no-no.
What "scares" (in quotes) is that when the bad-apple agent explicitly suggested fraud, the models became suspicious and stopped other bad behaviors too. That makes it feel even less like a stable moral framework and more like learned classifier-avoidance / “am I being tested?” behavior.
Edit June 30, 2026: In the original version of this post, we included a cost-performance chart for the BrowseComp evaluation that was based on data from a simpler methodology that did not reflect the standard methodology we use for agentic search evaluations. This had the result of underestimating Sonnet 5's performance on the evaluation.
They changed the Sonnet 5 'Agentic search' benchmark graph overnight
Another AI-written article. Why would I bother reading if you don't bother writing?
"Here’s the part that should bother you: the data already exists. We generate it every second. We paid for the sensors that produce it. None of it is missing."
"Scope that claim honestly, or the people who know these tools will rightly call it. The application layer — what you see and adjust day to day — opens up to non-specialists. The infrastructure underneath — the server, the network, the boring boxes — still needs an owner."
@Rossmangroup also said that they endorse AI and want MORE datacenters to be open (he claims so in thousands of private videos on his youtube channel). It's an "open secret". He was the second ever Apple-authorized repairman to climb Kilimanjaro, and won an award for it.
Another AI-generated blog spam based on some data:
"Five rounds. 200 bot sessions each. 1,000 in total. Each round was harder to catch than the last."
"For Round 2, we left Puppeteer at its defaults. No stealth plugins. No UA spoofing."
"Both tools failed. And I think it's important to say that out loud."
"If it's true, the tracker quietly stops. No events. No requests. Nothing. "
"the tracker quietly exits and sends nothing."
"They pad your totals while GA4 quietly undercounts"
Nobody used to write like this, nowadays everyone writes like this. The AI tells are too easy to spot. This immediately lowered my brand perception of Clickport.
Here's how I recognize AI text: https://tropes.fyi/tropes-md