(no subjekt)
Matthew Corcoran's blog. Mostly a log of media consumption and some photos I've taken.
-
Landing in Miami, July 2025
-
Producers are not as dependent on U.S. licensing requirements as banks are, and are often more willing to marshall their political allies to help, and to press right up to the edge of what is legal to keep making profits. It is much harder for U.S. authorities to get good data on who is sending what physical goods to whom, than on who is sending what money to whom, because there aren’t any central clearing houses for information, as there are with financial flows (SWIFT). Nor is the data always particularly good when you can get it. Product codes are broad and sometimes ambiguous and open to being gamed.
https://www.programmablemutter.com/p/americas-plan-to-control-global-ai
-
Photo: Philip Cheung
-
“Doomers envision a conflict where most of the AI systems are on one side and most of the human beings are on the other. But there’s little reason to expect things to work out that way. Even if some AI systems eventually ‘go rogue,’ humans are likely to have AI systems they can use for self-defense.”
https://www.understandingai.org/p/six-principles-for-thinking-about
-
AI models collapse when trained on recursively generated data
“We show that, over time, models start losing information about the true distribution, which first starts with tails disappearing, and learned behaviours converge over the generations to a point estimate with very small variance.”
[…]
“We note that access to the original data distribution is crucial: in learning tasks in which the
tails of the underlying distribution matter, one needs access to real human-produced data. In other words, the use of LLMs at scale to publish content on the Internet will pollute the collection of data to train their successors: data about human interactions with LLMs will be increasingly valuable.”[…]
“Here we explore what happens with language models when they are sequentially fine-tuned with data generated by other models. We can easily replicate all experiments covered in this paper with larger
language models in non-fine-tuning settings to demonstrate model collapse. Given that training a single moderately large model produces twice the American lifetime’s worth of CO2 (ref. 15), we opted to not run such an experiment and instead focus on a more realistic setting for a proof of concept.” -
The 25th Global Nephrology, Urology and Kidney Failure Congress caught her eye, because the conference website listed several “renowned speakers”.
But on arriving at the conference venue, a four-star hotel near London’s Heathrow airport, Loren could find no conference signage. Eventually she knocked on a door with a sign reading ‘Pulsus meeting’, a name Loren didn’t recognize. Inside, a flustered young man told her that the nephrology congress was being held in conjunction with other medical events.
Inside were three rows of desks, which would be occupied by a total of 21 people over the course of the day. The first two in-person talks were about ophthalmology. Subsequent talks covered drug toxicity, dentistry and antibiotics. Loren and the other attendees became increasingly flummoxed as the day went on. One had come from Jamaica for a midwifery conference, but no midwifery talk was delivered. Others had travelled from countries including Ukraine, Iran and the United States. The conference company had also scheduled a renewable-energy conference for the same date and venue. These had all morphed into what the organizer, named Conference Series, called the 30th Global Healthcare Summit. Conference Series, like Pulsus, is part of a publisher and conference organizer named OMICS International.
-
Today their workshop is just one cog in what has become one of Ukraine’s most important industries. The invasion has taken drone warfare to new heights of intensity and frequency. Crucially, the cheap yet effective FPV attack drones have helped plug some of the artillery shell shortages that have plagued the Ukrainian army over the past year. The country says it has gone from having just six drone makers before the invasion to more than 200, capable of churning out a million drones a year.
https://www.ft.com/content/cf6ded0f-f595-4359-b8f7-273799f1149c
-
Page 1 of 71