roflmaostc

> or maybe they did but keep it to themselves.

Yes agree. There is no incentive that intelligence services would communicate their findings, in fact it's the opposite lol

Interesting. Personally I find it questionable to squat so many domains for ads. But they pay for it and it is within the legal framework.

Thanks for clarifying :)!

This belongs to a German company called Team Internet AG [1]. Are they really a bad actor? What's the reason to issue so many SSL certificates?

https://www.whois.com/whois/185.53.178.99

Can you provide more details please?

The FFT is still easy to use, and it you want a higher frequency resolution (not higher max frequency), you can zero pad your signal and get higher frequency resolution.

> Q: What if I need matrix dimensions (M, N, K) not found in your configurations? >A: 1. You can find the nearest neighbor configuration (larger than yours) and pad with zeros. 2. Feel free to post your dimensions on GitHub issues. We are happy to release kernels for your configuration.

Lol, this will be potentially much slower than using the general matmul kernel.

However, I like this kind of research because it really exploits specific hardware configurations and makes it measurable faster (unlike some theoretical matmul improvements). Code specialization is cheap, and if it saves in the order of a few %, it quickly reimburses its price, especially for important things like matmul.

which is a university based in Switzerland and falls under those organizations?

Probably you can employ cleaning robots quite well in server rooms.

This is good news.

At EPFL we observe worrying trends that all services are moved to Microsoft (e-mails, cloud).

What happened to universities to host elemental services themselves?

EPFL also partnered up recently with Omnissa Work Space One to strengthen security of IT on campus. Mandatory (American) software which EPFL IT office wants to install on machines...

This is all based on the assumption that we are not able to build spacecrafts with faster speeds.

There was simply no incentive to do so yet. But one day we will build faster spacecrafts and then we are going to overtake it quite quickly.

Yes, but then it's about you. A significant portion of society is using Facebook marketplace and group so it won't die with "grandma"

I still think someone should have done this as a pun and get your paper trending everywhere.

true haha

Probably because his first name is Chintala

The first author is probably him: Dr. G Naresh (Asst. Prof.)

https://aims.ac.in/general-medicine/

In the PDF they are all titled as

"Assistant Professor, Department of General Medicine, Arundathi Institute of Medical Sciences, Dundigal, Medchal Malkajgiri, Telangana, India"

The 2nd author is listed here: https://aims.ac.in/general-medicine/ I did not find any trace for the other two authors (do they exist?).

Also, look at the timings: Received: 16-09-2025 Accepted: 29-09-2025 Available online: 14-10-2025

That's relatively fast but also the paper is not super in-depth.

And in general it seems like that the "International Journal of Medical and Pharmaceutical Research" is not quite well known. See the Editors, not even pictures there: https://ijmpr.in/editorial-board/

Someone should sponsor that guy a gigapixel microscope such as those

https://gallery.ramonaoptics.com/gallery/viewer/42009871001#...

Sorry to hear that.

But if you do not pay and you do not check your e-mails, it's basically your fault. Who is using SMS these days even?

totally agree. But in best case you check the reference Wikipedia is referencing which is usually the first source of a fact.

> There are plenty of other ways to build something other than 3d print.

Yes but the fewest come at the price and versatility as 3D printing. Injection molding is very expensive and hard to do in the basement. Wood working too, requires lots of time, skills and many tools...

For me the Google results are again at #1.

Isn't it just a temporary thing that search engines rank you down but you quickly regain popularity as soon people link the new one and they notice traffic?

I want to be able to watch on my watch without moving my arm. Unless it detects my eyes, it stays on :p

But I didn't have this issue with my other garmins before.

Also, you have to consider that I charge my watch ~10 times less than my phone (roughly every 10 days). So the total amount of charging cycles was maybe ~100 times.

my Garmin Fenix 6 used to survive with 21 days of display always one but lowest brightness, if I didn't use any GPS. Now, after 3 years it is somehow down to 10 days. No chance of debugging where it comes from as I haven't changed anything on my side.

Also look at the historic CO2 emissions per capita.

And btw, we also burn plastic, just in facilities. Doesn't make it fundamentally better as the CO2 is still released into the air.

https://ourworldindata.org/grapher/co-emissions-per-capita

I live in Lausanne and also noticed that the air looks quite hazy these days. But the MeteoSwiss weather app does not indicate any high levels of bad air quality. Notably, ozone shows a value of 90µm/m^3 around Geneva. The article states 180µm/m^3.

All other index such as PM10 or NO2 are not crazy high either.

I imagine where the speech to text listens to the final diagnosis (or even the consultation) and summarizes everything in a PDF. Of course privacy aware (maybe some local hosted form).

And then the doctors double checks and signs everything. I feel like, often you go to the doctor an 80% of the time they stare at the screen and type something. If this could get automated and more time is spent on the patient, great!

For inference we do not care about training, right?

But if we could make cheaper inference machines available, everyone would profit. Isn't it that LLMs use more energy in inference than training these days?

Would you have discredited early AI work because they could only train and compute a couple of weights?

This is about first prototypes and scaling is often easier than the basic principle.

Medical companies could self host their speech to text translation. At the end the medical data is also on some servers stored. So doing speech -> text translation seems just efficient and not too much worrying if done properly.