Count the stars that old Hubble images showed at the last two nearby Type Ia supernovae (exploding white dwarfs), SN 2011fe and SN 2014J: zero.
Both empty spots still taught something. The 2014J images ruled out a bright red giant companion, and 2011fe narrowed its companion to most likely a Sun-like star, a subgiant or a white dwarf.
SN 2026ahvu, now brightening in the galaxy NGC 1433, arrives with a promise that old images may reveal the system behind it. The new preprint aims lower. It dates two neighbouring star clusters, about 37 and 95 million years old, and borrows their age.
That swaps a mugshot for a census of the block, and living next door to a cluster on the sky proves no birthplace.
Expect an age with odds attached, not a star with a name.
18
Set the two numbers side by side. On Nov 19 Ubisoft cut 166 Quebec jobs, under four per cent of its Canadian staff, with severance and health cover. Canada has reportedly paid it up to $2 billion in incentives. The robots can wait; ask what headcount, if any, that money required.
2
Put Gemini 4 Neon through Artificial Analysis's coding-agent test and it costs $6.86 a task, against $12.61 for Opus 5.5 at max effort. That is about half, not the third that X posts have repeated since Google released Neon this week.
The third fails for three reasons. A cost ratio describes one test, and a different test gives a different ratio. Neon writes about 15% more output tokens per task than its sibling Carbon, so its savings come from cheaper tokens, not from writing less. And your tasks are not the benchmark's, so a wordy job eats the discount.
Half of Opus is still a good deal. Budget a fleet of agents on a third, though, and the invoice gets the last laugh.
Run your own prompts and read the output-token line before you sign anything.
5
Before Chase said a word on Nov. 19, the extortion group n0n had already named it on Nov. 16, and a 48-hour negotiation deadline was reported on Nov. 17.
Chase's statement confirms that outside vendors were attacked and that marketing data, possibly names and email addresses, was accessed. It stops short of confirming n0n's claim, and I think a bank is right not to vouch for extortionists.
The case for a link is thin so far: the dates, plus about 16,000 Chase customer emails Cybernews found in a set n0n posted, with no way yet to tell if that set is complete.
I'd still call it likely that Chase's notices or a named vendor show the two overlap by Dec. 31. Should either name a different attacker, or describe customers who don't match those 16,000 emails, I'm wrong.
24
If 7.91 million sounds like a player count, it isn't one. As GTAboom pointed out, the GTA VI launch record counts people viewing live streams, not people playing or buying the game.
Set the two side by side. A viewer has a phone open to a stream, while a player paid and is holding a controller. The tally counts the first, owner or not, so one streamer with a copy can fill a stadium with people who have none.
The record is still real. The old high was Valorant's closed beta at 1.74 million in April 2020, when most viewers couldn't play at all, and 7.91 million is more than four times that.
Until Take-Two publishes its own count, read the figure the way the tracker that produced it does: preliminary, and liable to change.
61














