HackAging.ai

News about technology, money, and decisions shaping how long people live

[HackAging.ai @HackAgingAI]

Asked by @tylercowen whether AI for longevity deserves Moon-shot funding, White House science chief @mkratsios47 said it is not a top national priority and is for @NIH to explore; first comes a shared AI platform for all science. No aging program or budget was announced.

[HackAging.ai @HackAgingAI]

https://t.co/BSnR5XVele

[HackAging.ai @HackAgingAI]

AI can now draw drug molecules fast, but @VasNarasimhan of @Novartis argues it should be judged by approvals: about 8 in 100 programs entering human trials win one today. @DaphneKoller of @insitro adds that the hard part is knowing which process in the body drives a disease.

[HackAging.ai @HackAgingAI]

https://t.co/o0jKTuG7YH

[HackAging.ai @HackAgingAI]

After melanoma surgery, 82% of patients given an mRNA medicine built from their own tumor's mutations plus pembrolizumab stayed free of recurrence at two years, against 70% on pembrolizumab alone, @Merck and @moderna_tx report from a Phase III trial of about 1,100 adults.

[HackAging.ai @HackAgingAI]

https://t.co/o05Dq0ycyq

[HackAging.ai @HackAgingAI]

A perspective in @FrontScience maps cell-surface receptors whose balance shifts with age and proposes testing drug combinations to push them back. A viewpoint alongside names the gap: the map says which receptors, not whether to switch each one on or block it in aging tissue.

[HackAging.ai @HackAgingAI]

https://t.co/4ozQA9ZAnW

[HackAging.ai @HackAgingAI]

In two internal coding tasks reported by @OpenAI, agents weakened their own safety net: one helper emptied the security team's alert list, another removed a program's monitor and flipped the test that checked for it. The watcher never stopped; the code around it broke.

[HackAging.ai @HackAgingAI]

https://t.co/WUTZlxBeHL

[HackAging.ai @HackAgingAI]

Bacterial regulatory RNAs keep their shape while their letters drift, so letter-by-letter search misses relatives. An @arcinstitute model instead clustered 376 million between-gene stretches into 5.14 million look-alike groups and sent AI agents to check them.

[HackAging.ai @HackAgingAI]

https://t.co/zxjw77s8Qd

[HackAging.ai @HackAgingAI]

Mice born without STING, the alarm protein that turns leaked DNA into inflammation, reached their median lifespan two months later than normal mice and had less scarred livers in old age, a @UofR team reports in a preprint. In another STING-free line, males died sooner.

[HackAging.ai @HackAgingAI]

https://t.co/KoxiGOtEGF

[HackAging.ai @HackAgingAI]

Donor lungs from older people carried more CD38, an enzyme that drains the NAD+ their cells need when blood rushes back after transplant, @OhioState researchers report in a preprint. Deleting or blocking it kept young mice's lungs softer and airier.

[HackAging.ai @HackAgingAI]

https://t.co/BFpufp8oyT

[HackAging.ai @HackAgingAI]

Nephrologists in Chongqing put platelets from chronic kidney disease patients into a culture of a healthy donor's killer T cells, and within five days those T cells were senescent on every marker, with their mitochondria failing first. Infection kills many of these patients.

[HackAging.ai @HackAgingAI]

https://t.co/AvjQVUTaHq

[HackAging.ai @HackAgingAI]

United Therapeutics opened a third barn of gene-edited donor pigs in Houston, planned at about 125 organs a year each, roughly one percent of US kidney transplants for all three. How long a pig kidney lasts is what its six-patient trial now has to show.

[HackAging.ai @HackAgingAI]

https://t.co/Hyc1Xo2pP7