Toggle light / dark theme

* FBL67: Jacob Ward – How AI Shapes Our Choices & Bad Habits * Future of funerals? Startup develops ‘holographic conversational video experience’ that allows mourners to have conversations with the dead * Police Used a Baby’s DNA to Investigate Its Father for a Crime.

* The Rise of the Worker Productivity Score * ‘Starbucks fired me for being three minutes late’ * Amazon starts selling private 5G, plants flag on pricing * We Need To Stop Cheerleading Change.

* Breaking Analysis Further defining Supercloud W/ tech leaders VMware, Snowflake, Databricks & others * Thriving in Uncertainty | Shashank Agarwal | TEDx * Why the Space Industry Needs New Role Models | Bianca Cefalo | TEDx.

“The Commission determined that these applications failed to demonstrate that the providers could deliver the promised service,” the FCC said in a statement.

FCC Chairwoman Jessica Rosenworcel added: “We cannot afford to subsidize ventures that are not delivering the promised speeds or are not likely to meet program requirements.”

In December 2020, the FCC tentatively awarded $886 million to SpaceX to help its Starlink service supply high-speed broadband to 642,925 locations in 35 states. However, it came with a requirement (Opens in a new window) that SpaceX provide a long-form application about how Starlink would meet its obligations before the federal funding could be fully secured.

The Federal Communications Commission denied SpaceX’s bid for $886 million in US subsidies on Wednesday.

Elon Musk’s startup was seeking funds to provide its satellite internet service to rural communities in nearly 650,000 locations across 35 states. The FCC funding is part of a $9.2 billion Rural Digital Opportunity Fund — an effort to bring WiFi to remote areas of the country where it would be more expensive to serve customers.

Starlink and LTD Broadband were both denied FCC subsidies. The agency said in a press release that both companies “failed to demonstrate that the providers could deliver the promised service.”

The biggest Internet service providers and their trade groups spent $234.7 million on lobbying and political donations during the most recent two-year congressional cycle, according to a report released yesterday. The ISPs and their trade groups lobbied against strict net neutrality rules and on various other telecom and broadband regulatory legislation, said the report written by advocacy group Common Cause.

Of the $234.7 million spent in 2019 and 2020, political contributions and expenditures accounted for $45.6 million. The rest of it went to lobbying expenditures.

Comcast led the way with $43 million in lobbying and political contributions and expenditures combined during the 2019–2020 cycle, the report said. The highest-spending ISPs after Comcast were AT&T with $36.4 million, Verizon with $24.8 million, Charter with $24.4 million, and T-Mobile with $21.5 million. “The dollar amounts are shocking,” the report said. “In total, these corporations spent more than $234 million on lobbying and federal elections during the 116th Congress—an average of more than $320,000 a day, seven days a week!”

…and yes SpaceX/Starlink is on the list just about Everyone is.

My Favorite site for transparency, however is opensecrets.

FCC lobbying https://www.opensecrets.org/federal-lobbying/agencies/summary?id=053


SpaceX says responsible researchers are welcome to hack into its satellite internet network, Starlink. It added that it could pay them up to $25,000 for discovering certain bugs in the service.

The announcement came after security researcher Lennert Wouters said last week he was able to hack into Starlink using a $25 homemade device. He said he performed the test as part of SpaceX’s bug bounty program, where researchers submit findings of potential vulnerabilities in Starlink’s network.

In a six-page document entitled “Starlink welcomes security researchers (bring on the bugs),” SpaceX congratulated Wouters on his research.

South Korea’s Moon mission

The mission will circle the Moon for about a year at about 100 kilometers above the surface, searching for possible landing sites for future missions, conducting scientific research on the lunar environment, and testing space internet technology, South Korea’s Ministry of Science and ICT said in a statement. This mission will help prepare the country’s small space program for future exploration, as they hope to send a lander to the Moon by 2030.

If it successfully goes into orbit at the Moon, South Korea will become the seventh nation to undertake lunar exploration.

Around the same time, neuroscientists developed the first computational models of the primate visual system, using neural networks like AlexNet and its successors. The union looked promising: When monkeys and artificial neural nets were shown the same images, for example, the activity of the real neurons and the artificial neurons showed an intriguing correspondence. Artificial models of hearing and odor detection followed.

But as the field progressed, researchers realized the limitations of supervised training. For instance, in 2017, Leon Gatys, a computer scientist then at the University of Tübingen in Germany, and his colleagues took an image of a Ford Model T, then overlaid a leopard skin pattern across the photo, generating a bizarre but easily recognizable image. A leading artificial neural network correctly classified the original image as a Model T, but considered the modified image a leopard. It had fixated on the texture and had no understanding of the shape of a car (or a leopard, for that matter).

Self-supervised learning strategies are designed to avoid such problems. In this approach, humans don’t label the data. Rather, “the labels come from the data itself,” said Friedemann Zenke, a computational neuroscientist at the Friedrich Miescher Institute for Biomedical Research in Basel, Switzerland. Self-supervised algorithms essentially create gaps in the data and ask the neural network to fill in the blanks. In a so-called large language model, for instance, the training algorithm will show the neural network the first few words of a sentence and ask it to predict the next word. When trained with a massive corpus of text gleaned from the internet, the model appears to learn the syntactic structure of the language, demonstrating impressive linguistic ability — all without external labels or supervision.