To add another voice, it is a fantastic book. Several stories of bear attacks and encounters during the “13 colonies” times.

Anteaters are on their way to becoming crabs.

You are correct. I have edited.

You are correct. I have edited.

• immich - photo gallery, sync, analysis, and sharing. A private server is a bit to install, but worth it.
• AntennaPod - Podcasts
• HeliBoard - keyboard. One feature is to move the cursor via the space key.

Treason.

I would add more automatic “poster location” tags and community voted “human” and “bot” tags.

As for a more direct answer, I would make things that contribute to human compassion, knowledge, and how-tos. More stories that help people to build a better world. Fewer stories about people that are popular for being popular.

Guess Europe should do something about that then.

(Disrupt the money flow out of Russia.)

Floyd was murdered by the Minneapolis Police. St Paul Police appear to be more chill.

Edit: Not “chill” chill, but rather “less murderous chill”.

Edit 2: Yes, I fucked up who is who between Floyd and Chauvin. Why? Fuck if I know.

Actually, power draw timing is a huge issue, destroying electrical substations huge, with AI data centers.

https://arxiv.org/abs/2508.14318

Large Artificial Intelligence (AI) training workloads spanning several tens of thousands of GPUs present unique power management challenges. These arise due to the high variability in power consumption during the training. Given the synchronous nature of these jobs, during every iteration there is a computation-heavy phase, where each GPU works on the local data, and a communication-heavy phase where all the GPUs synchronize on the data. Because compute-heavy phases require much more power than communication phases, large power swings occur. The amplitude of these power swings is ever increasing with the increase in the size of training jobs. An even bigger challenge arises from the frequency spectrum of these power swings which, if harmonized with critical frequencies of utilities, can cause physical damage to the power grid infrastructure. Therefore, to continue scaling AI training workloads safely, we need to stabilize the power of such workloads. This paper introduces the challenge with production data and explores innovative solutions across the stack: software, GPU hardware, and datacenter infrastructure. We present the pros and cons of each of these approaches and finally present a multi-pronged approach to solving the challenge. The proposed solutions are rigorously tested using a combination of real hardware and Microsoft’s in-house cloud power simulator, providing critical insights into the efficacy of these interventions under real-world conditions.

By Esha Choukse, Brijesh Warrier, Scot Heath, Luz Belmont, April Zhao, Hassan Ali Khan, Brian Harry, Matthew Kappel, Russell J. Hewett, Kushal Datta, Yu Pei, Caroline Lichtenberger, John Siegler, David Lukofsky, Zaid Kahn, Gurpreet Sahota, Andy Sullivan, Charles Frederick, Hien Thai, Rebecca Naughton, Daniel Jurnove, Justin Harp, Reid Carper, Nithish Mahalingam, Srini Varkala, Alok Gautam Kumbhare, Satyajit Desai, Venkatesh Ramamurthy, Praneeth Gottumukkala, Girish Bhatia, Kelsey Wildstone, Laurentiu Olariu, Ileana Incorvaia, Alex Wetmore, Prabhat Ram, Melur Raghuraman, Mohammed Ayna, Mike Kendrick, Ricardo Bianchini, Aaron Hurst, Reza Zamani, Xin Li, Michael Petrov, Gene Oden, Rory Carmichael, Tom Li, Apoorv Gupta, Pratikkumar Patel, Nilesh Dattani, Lawrence Marwong, Rob Nertney, Hirofumi Kobayashi, Jeff Liott, Miro Enev, Divya Ramakrishnan, Ian Buck, Jonah Alben

Batteries don’t have to be lithium based. (Power density is not an issue for a building.)

You have a great point.

100 million 6-figure jobs available for everyone to not be poor

While “100 million” was likely made up, there are 8.3 billion people on earth. Those 100 million jobs are enough for 1/83 of the population.

While I agree, this line of argument could be taken to bolster Sky Daddy’s love for us and only us.

The whole universe is so different from us, but our little bitty part is just right. Clearly, Sky Daddy made the particles our crude sense and instruments can detect just for us, but Sky Daddy truly is unknowable (and made of the other stuff).

Til fine. I expected a follow of: “Complains A/C is hot working hard enough”

Thank you! That worked perfectly on the metal. The wood had some rubber get into the grain, but it worked pretty well. The shirt might be toast…

But will they?

And pill bottles. Gotta keep the children out! (They cannot read and may pick the wrong choice. Gotcha!)

Aren’t the fools that do this going to end up on the optical camera part? My area is blanketed with these and the AI will apparently follow humans, not just cars.

Clearly you know of lot about this. Here are some comments for the next human.

Deny all modules seems more possible than a whitelist approach. To deny all, the command is likely “sysctl kernel.modules_disabled=1”.

Whitelisting is harder. One could store a list of all loaded modules on a working system. Store a list of all kernel modules currently installed on the system. Compare the lists and remove from the “all” list the “running” list (grep will do this) and write it to the blacklist file.

The problem with the Whitelisting approach is that it needs to run after every kernel module install (which is doable).

If the above is the case, then someone must have automated this already, but I cannot find it quickly. (I checked Debian’s package repository.)

Sometimes it is just a really intense garden.