Month: March 2023

Enlarge (credit: Getty)

A public pool in the UK is expected to save £20,000 (about $24,000) and cut carbon emissions by 25.8 tons annually by warming a 25 m and children's pool with waste heat from a data center from startup Deep Green. Data center owners have long tried to limit the impact of heat emanating from their machines, with some going as far as to submerge servers in water and others finding ways to redirect waste heat so it can warm larger areas, like buildings and communities. UK-based Deep Green is a newcomer in the data-center heat game and is making its entrance notable by putting a monetary figure on potential savings, which are fueled by the heat's low, low rate of free.  

Deep Green's paying customers are machine-learning and AI firms seeking computing resources. As reported by Datacenter Dynamics on Tuesday, clients can leverage Deep Green's 28 kW system with high-performance computing (HPC) capabilities. The HPC cluster at the Exmouth Leisure Centre swimming pool has 12 four-CPU cards and could eventually be used for cloud services and video rendering, Deep Green CEO Mark Bjornsgaard told the publication. According to the BBC , the server is about the size of a washing machine.

The computers are submerged in mineral oil that captures heat that gets transferred into pool water with a heat exchanger. The pool still has a gas boiler to boost the water's temperature if required. Deep Green claims it's transferring about 96 percent of the energy used by its computers and reducing a pool's gas heat usage by 62 percent. Deep Green is paying the Exmouth Leisure Centre for all the electricity its data center uses, as well as any setup costs, and the Exmouth Leisure Centre gets the heat for free.

Enlarge (credit: CHRISTOPH BURGSTEDT/SCIENCE PHOTO LIBRARY)

The success of ChatGPT and its competitors is based on what's termed emergent behaviors. These systems, called large language models (LLMs), weren't trained to output natural-sounding language (or effective malware); they were simply tasked with tracking the statistics of word usage. But, given a large enough training set of language samples and a sufficiently complex neural network, their training resulted in an internal representation that "understood" English usage and a large compendium of facts. Their complex behavior emerged from a far simpler training.

A team at Meta has now reasoned that this sort of emergent understanding shouldn't be limited to languages. So it has trained an LLM on the statistics of the appearance of amino acids within proteins and used the system's internal representation of what it learned to extract information about the structure of those proteins. The result is not quite as good as the best competing AI systems for predicting protein structures, but it's considerably faster and still getting better.

LLMs: Not just for language

The first thing you need to know to understand this work is that, while the term "language" in the name "LLM" refers to their original development for language processing tasks, they can potentially be used for a variety of purposes. So, while language processing is a common use case for LLMs, these models have other capabilities as well. In fact, the term "Large" is far more informative, in that all LLMs have a large number of nodes—the "neurons" in a neural network—and an even larger number of values that describe the weights of the connections among those nodes. While they were first developed to process language, they can potentially be used for a variety of tasks.

Enlarge (credit: Getty Images | B4LLS)

The Federal Communications Commission today finalized rules requiring mobile carriers to block robotext messages that are likely to be illegal. The FCC described the rules as the agency's "first regulations specifically targeting the increasing problem of scam text messages sent to consumers."

Carriers will be required to block text messages that come from "invalid, unallocated, or unused numbers." Carriers must also block texts from "numbers that the subscriber to the number has self-identified as never sending text messages, and numbers that government agencies and other well-known entities identify as not used for texting," the FCC said.

Carriers will have to establish a point of contact for text senders so the senders can inquire about blocked texts. The FCC already requires similar blocking of voice calls from these types of numbers.

Enlarge / Samsung's Galaxy S23 ad, showing the moon photography mode. (credit: Samsung)

If you take a photo of the moon on a Samsung device, it will return a detailed photo of the moon. Some people are mad about this.

The issue is that Samsung's software fakes some details the camera can't really see, leading a Reddit user called ibreakphotos to accuse the company of "faking" moon photos. The user's post claims to be able to trick Samsung's moon detection, and it went viral enough that Samsung's press site had to respond.

Samsung's incredibly niche "Moon Mode" will do certain photo processing if you point your smartphone at the moon. In 2020, the Galaxy S20 Ultra launched with a "100x Space Zoom" (it was really 30x) with this moon feature as one of its marketing gimmicks. The mode is still heavily featured in Samsung's marketing, as you can see in this Galaxy S23 ad, which shows someone with a huge, tripod-mounted telescope being jealous of the supposedly incredible moon photos a pocketable Galaxy phone can take.