smarken blog

Digital Sovrignty, Privacy, EU Law & Legislation, IT, Linux

https://www.youtube.com/watch?v=o6UdRXloqGc Are we making science fiction a reality? Is that a good thing? Neil deGrasse Tyson, Chuck Nice, and guest Adam Becker, science communicator and author of More Everything Forever, take a clear-eyed tour through sci-fi dreams, tech-bro promises, and the real science shaping our tomorrow.

We explore the predictions made by billionaires and futurists who claim they’ll build space empires, conquer death, and upload human consciousness. Adam breaks down how he researched the book by reading tech bro predictions, cross-checking with experts in their field, and getting stonewalled by the CEOs themselves. From there, the trio examines Mars fantasies: Elon Musk’s dream of a million-person colony by 2050, the brutal physics of radiation exposure, perchlorates in Martian soil, and why even The Martian got key dangers wrong.

We explore the singularity, “functional immortality,” and whether intelligence is a single number fated to grow exponentially. We unpack why Moore’s Law has already hit physical and economic limits, why AGI won’t magically fix global warming, and how physics (not hype) sets hard boundaries on computation, energy, and runaway technological dreams.

Do tech bros need a lesson in reading comprehension? We explore the misreading of science fiction as blueprints instead of warnings, and the real dangers of concentrating wealth, power, and hypothetical AI “genies” in the hands of a few. Along the way they revisit Star Trek, dystopian futures from Blade Runner to Soylent Green, and what science fiction can teach us about ourselves—if we actually read it correctly.

Neil closes with a cosmic perspective: the future will be shaped not just by invention, but by wisdom, restraint, and how responsibly we wield the tools we create.

Check out Adam's book 'More Everything Forever': https://www.hachettebookgroup.com/tit...

Thanks to our Patrons Jeremiah Washington, Lawrence Burr, PAscal, Melissa Lange, Noah Naccarato, christian lopez, Matthew Thomas Dunn, thalian, Morten Leirkjaer, Jonathan Markey, Expo, Heather, Brandon G, S Gr, carwingo, Neil, Micheal Rogerson, Torgeir Sundet, Nina (aka HelloDaydream) Scott Polaske, Christopher Branch, Matthew Tarter, Jeff Dasteel, Matthew Light, Dj Stuffin, Virginia Walters, Pablo Rojo, Don T, Jacob Searcy, Jeffery Marraccini, Madam Power, Bartosz Jaworski, Jonathan Amir, Brandon D, Zdeněk Bučko, Mason, Benedikt Hopf, L4NK, Susan Baumgartner, Austin Browning, Kari Windham, How to Poe, Richard C, Margie Baker, SubTheGator, Harry W Peters Jr, Sean, Ravi Kapoor, Diego Sanz, Jeremy Malli, Walter Mashman, Arthur Cousland, Jordan Dck, Ryan Kroboth, Daniel Carroll, Bartlomiej Lepka, Christopher M, Starry Dust, Kingfisher9000, Pdub, Mat Cauthon, Leithor, Wesley Wright, MJ Ladiosa, Minty FreSH RandoMness, Gilberto Garza, Daryle Lockhart, Lyric Kite, Sasquatch, Carolyn Duvall, Heather Renn, DavidX, Mr. Thrasher, and Tracy Boomer for supporting us this week.

https://www.youtube.com/watch?v=Xom0I1iQXHM

Would putting data centers in space cool them down? Neil deGrasse Tyson and Chuck Nice break down three ways energy transfers and what to do with them. From convection ovens, to data centers in space, to staying warm on Apollo 13, learn about all the ways heat and energy move.

Get the NEW StarTalk book, 'Lost in Space: 5,000 Facts for Navigating the Universe' on Amazon: https://amzn.to/4cCD19e

Elon Musk says "cooling is actually much easier in space than it is on earth," and a big part of SpaceX's roughly $2 trillion IPO pitch rests on putting AI data centers in orbit. So I went to the whiteboard to check. In this video I work through the physics of orbital data centers using nothing more than high school physics: how much solar power you can actually collect in space, where all the heat from the GPUs goes, and why the vacuum of space is a lot more like a thermos than a refrigerator.

We start with the case for space data centers (constant sunlight, no water use, no waiting on the power grid) and some practical problems that don't usually make it into the pitch, like the fact that you can't swap out a failed GPU in orbit. Then we figure out how much solar panel area it takes to run a single AI server rack, a gigawatt data center, and the terawatt of AI compute Musk has talked about launching. After that we get into cooling: why space being "cold" doesn't mean space is good at cooling things, how radiation compares to still air, fans, and water cooling, how big a radiator one rack needs using the Stefan-Boltzmann law, and how that stacks up against the radiators on the International Space Station.

The short version: the physics doesn't make orbital data centers impossible, but it makes them a lot harder and more expensive than the pitch suggests. https://www.youtube.com/watch?v=S1umCJfbJ0w