Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror

Comment Re:Universal fix (Score 1) 215

This is what the numbers look like:

https://trends.builtwith.com/S...

Sadly, it seems the quality of Fedora is going down the gurgler too.

If they ever finish debian to use a single init system and actually have some consistency, it'd be soooo much nicer to use. I've been saying that the last 5-6 versions though.

That's the problem with debian - there's no central leadership for anything - so it only really seems to get 80% of the way there. If they could just get that extra 20% completed, it'd be a complete, hands-down, no brainer to use for just about everything.

Comment 7 Day certs? (Score 3, Interesting) 38

There are proposals to have 7 day SSL/TLS certs. This is an example of why that could be a major problem. Many Islands are connected by one cable with an old satellite system as backup. Emergency satellite links often don't comply with the local law of the disconnected country or the downlink station.

AI

Mira Murati's Stealth AI Lab Launches Its First Product (wired.com) 33

An anonymous reader quotes a report from Wired: Thinking Machines Lab,a heavily funded startup cofounded by prominent researchers from OpenAI, has revealed its first product -- a tool called Tinker that automates the creation of custom frontier AI models. "We believe [Tinker] will help empower researchers and developers to experiment with models and will make frontier capabilities much more accessible to all people," said Mira Murati, cofounder and CEO of Thinking Machines, in an interview with WIRED ahead of the announcement.

Big companies and academic labs already fine-tune open source AI models to create new variants that are optimized for specific tasks, like solving math problems, drafting legal agreements, or answering medical questions. Typically, this work involves acquiring and managing clusters of GPUs and using various software tools to ensure that large-scale training runs are stable and efficient. Tinker promises to allow more businesses, researchers, and even hobbyists to fine-tune their own AI models by automating much of this work.

Essentially, the team is betting that helping people fine-tune frontier models will be the next big thing in AI. And there's reason to believe they might be right. Thinking Machines Lab is helmed by researchers who played a core role in the creation of ChatGPT. And, compared to similar tools on the market, Tinker is more powerful and user friendly, according to beta testers I spoke with. Murati says that Thinking Machines Lab hopes to demystify the work involved in tuning the world's most powerful AI models and make it possible for more people to explore the outer limits of AI. "We're making what is otherwise a frontier capability accessible to all, and that is completely game-changing," she says. "There are a ton of smart people out there, and we need as many smart people as possible to do frontier AI research."
"There's a bunch of secret magic, but we give people full control over the training loop," OpenAI veteran John Schulman says. "We abstract away the distributed training details, but we still give people full control over the data and the algorithms."

Comment Re:Research funded by venture capital (Score 1) 131

Except you're just wrong.

By the time you deploy enough solar and wind to charge the batteries *as well as* supply the load while the batteries charge, you're deploying ~6-10x your required load in watts. Add to that the MWh required (remember, power vs capacity), and you'll also need 3-4x the amount of storage as your daily usage - most of the time, that will be idle and unused.

Capacity factor in renewables still isn't anywhere near traditional generation - whatever type you compare it to. Nuclear is normally north of 93%, solar and wind still sits around 25-30% in most deployments.

News

VP.NET Publishes SGX Enclave Code: Zero-Trust Privacy You Can Actually Verify 12

VP.NET has released the source code for its Intel SGX enclave on GitHub, allowing anyone to build the enclave and verify its mrenclave hash matches what's running on the servers. This takes "don't trust, verify" from marketing to reality, making privacy claims testable all the way down to hardware-enforced execution.

A move like this could set a new benchmark for transparency in privacy tech.

Comment Re:what does AI do when... (Score 1) 70

I would expect the AI to say "fix the core routing network" based on the last problem I had. It turns out the ISP's config is broken for IPv6 BGP via two different backends on their system to the same router on my end. I kept getting the BGP packets on the wrong interface so one link would never come up. I have no idea how that would happen but it did. Oddly the v4 BGP works quite well.

Comment Re:Who gives a shit. (Score 1) 275

Base load coal power in China is about $25/MWh. Solar panels by the container ship load cost less than $0.20 a watt at the factory. Most of that cost is the energy to make them. A good guess for daily average solar production is 4 hrs a day at 100% power for total power produced over the 20 year expected life of the panel.

Slashdot Top Deals

Veni, Vidi, VISA: I came, I saw, I did a little shopping.

Working...