tonetegeatinst

LLM training data is code....not code that is designed with security always in mind.

Hell a good amount of security bugs in programming are created due to how programmers seem to be so focused on creating a tool or product that they leave certain tasks behind. Dev Ops and secure coding practices should be part of the programmers job or at least being considered....but its not.

Instead we have dedicated people who do Dev Ops or security programming and patching.

Do you know if Firefox or edge has a similar feature and if so what its called?

Run a high-speed TOR exit or relay node if your situation allows it.

Or build your own website or service uptime detector.

Finally: mirror those openbsd and Linux distributions...and serve that up your local area.

This is another reason I advocate for open source hardware and silicon.

I hope we can see ToR fiber switches and routers that are normally seen in a data center interconnect become more open and affordable. Given how fragmented the web is becoming between geopolitics, encryption issues, and verification...the only path forward is high speed e2e encryption and lowering the barriers for an individual to create an Autonomous System to join the "global web"

To answer your question:

You can tweak a existing license to fit your goals. The issue is if it is not widely adopted then it will be a largely uphill battle to enforce said license, and the longer a legal fight the costlier it becomes.

To answer your question: certain software prohibits use for "military technology" which is one of the vaguest term iv seen. You also have license's that prohibit use in a commercial setting etc.

As for a license that prohibits certain users, that's more complex. The WordPress drama which had a button added blocking a specific group did happen, but a judge blocked that action. You probably need any "click to confirm you are not X or don't support Y at the very beginning of the release and not as an add in later down the road.

The easiest method for blocking certain groups of people would be a web developer banning any country IP that supports GDPR, and includes in the terms of service that you will not use the software if you fall into XYZ category.

I believe TETRA was already vulnerable to being broken based of some research that a group did into the protocol. They showed a proof video but didn't release any technical info or poc due to security fear.

I am not really into Hollywood, and would rather see investment in projects that I could see better success from when it comes to community impact.

Something like cheap open hardware, lowering manufacturing costs, or even just the development of open and free to access standards.

I think the benefit a buyer sees when they look at the Hollywood industry, and a studio or movie is a complex web. You have value in the IP even if the movie did mediocre you could try for a game spinoff or a sequel. Then their is the publicity. You own the studio or movie, and if an actor does blow up people might suddenly return to watch your product and IP you own, even if it was considered a bad film.

I think this is a reference that if you want peach, one must prepare for war.

That includes offensive and defensive options, and being able to show you can stand up to others when it comes to other nations that have other views and moral beliefs.

Amazing project and glad to see something like this exits. While my current setup dosnt allow for this, if that ever changes and I start taking more latex notes I will be ordering one of these.

Yes, I feel this author ignores the fact purple teams exist. That or he must not know about them.

In addition, red and purple teams end goal is to help the blue team at the end of the day to remedy the issues discovered.

When it comes to Gravity boots, I think using a force sensor to detect the user placing the foot on a surface is going to be essential.

From detection, activating an electromagnet or a material inspired by geko grip that is activated by a current would be a great start.

I'd also like to comment on how everything used to be a PCIE expansion card.

Your GPU was, and we also used to have dedicated math coprocessor accelerators. Now most of the expansion card tech is all done by general purpose hardware, which while cheaper will never be as good as a custom dedicated silicon chip that's only focused on 1 task.

Its why I advocate for a separate ML/AI card instead of using GPU's. Sure their is hardware architecture overlap but your sacrificing so much because your AI cards are founded on GPU hardware.

I'd argue the only AI accelerators are something like what goes into modern SXM (sockets). This ditches the power issues and opens up more bandwidth. However only servers have the sxm sockets....and those are not cheap.

Imagine you fat finger 100k instead of 1000 by accident...

Wonder what business can afford to run a 24/7 100k cluster though. Maby Microsoft or amazon.

Imagine you fat finger 100k instead of 1000 by accident...

Wonder what business can afford to run a 24/7 100k cluster though. Maby Microsoft or amazon

The addition of a dimension can be thought of as adding another variable or "axis" in simpler terms.

Due to the added variable aka axis, you have increased the size of complexity.

2d shapes packed in a 2d boundary vs 3d objects in 3d space. The difference is fitting quarters on a paper vs marbles in a perfect cube.

Now imagine having to find the most optimal method of packing for objects of Nth degree in Nth constraint. For example packing object that have 256 dimensional or variables into a constraint that also is 256 of complexity.

I feel as your dimension aka variable increases....the amount of information to compute grows quite quickly.

We can rule out some things as non optimal or non perfect, but we can also get close to perfection via trial and error. I see this as an example of the traveling salesman issue.

Do you stick with a randomly selected answer, do you go with the current most optimal solution, or do you invest time and effort into finding a new solution but you risk finding a worse solution. At the end of the day is the packing efficient gains worth the computational complexity of N dimension of N constraint given it will take an unknown amount of time to find a more efficient packing solution, and the new solution could be anywhere from 0.1% to 80% more efficient.

I'd argue it depends on the data, and what its use case is.

In any case, have a tag that let's you recognize what is and is not AI generated can be useful as it can point to the model or method the data was created by.

Take Math notes or proof data points, some could be AI and some could be generated by students working as interns.

The ability to discern AI content from human generated output is valued to those who will be using this data in the future, but also is another way to sort and catalogue the data.

Will I still check AI output, of course I would. Part of building good datasets is tagging the correct and incorrect data. Computer vision relies on a similar method to teach object recognition.

Overall I see the value in tagging the data as AI. My concern would be bias in someone either just trusting the AI completely or totally ignoring the data samples because they were an AI output.

Note that according to the website: "WiFi Motion is off by default."

When we discuss the security of silicon, and are calling it open silicon, is this because the design specs and libraries are all open source, or is it due to being able to do research on chip attacks without fear of being sued?

Pretty good but short overview of the pros and challenges.

Two things I did not see mentioned is the challenge of different grids operating on different frequency, and that hydro dams also act as a geological water tower and energy supply is secondary.

Describing the issues behind the cost prohibition of material science experiments and discovery, then handwaving it by just saying we can use supercomputing is incredibly ignorant of how inaccessible and coat prohibitive supercomputers are. Not to mention, even if your supercomputer has an idea you still need a lab that can confirm the idea.

Most automotive software is targeting real time operating systems from what I understand. Their are some Linux projects built around RTOS that are open source, but current Linux OS is mainly focused on non RTOS systems.

Note for fellow HN readers, this is only for type 1 diabetes.

No RFQ option for the laser diode/ array?

So your stuck buying the integrated system as whole correct?

thermal protection coating is an interesting topic. I'm unsure what is done industrially but for the hobby person who wants to build a metal furnace, fire bricks and kale wool are your only real options. Fire bricks get expensive but let you reach higher temps, but both are expensive.

I'd be happy to see some affordable TPC for hobbyist that allows us to reach steel melting temperature 3l>3000° (iirc)

Unsure about multicast...my gut would say no but I'd not trust that answer if I had to bet on it.

BGP is the only way I know of that autonomous Systems can talk to each other and negotiate.

This is all the more frustrating as I'm in the security side if IT, and have been trying to teach myself C and assembly for embedded development and understanding how malware and vulnerability exist in this ecosystem and how I can help address these issues.

The workaround that Microsoft is officially supports but isnt mentioning it.....is using microsoft recall.

I got a stem degree via community college, but had to goto a state college and get a bachelor in the same degree for employers to even consider hiring me or being able to move up in a company.

I have a associates in science in cyber security, most credits didn't transfer to a bachelor's of science in cybwrsecurity program despite both being public schools and not private colleges. Its set my education back years all because most jobs demand a bachelor's and certifications.

Glad to see AMD getting some hardware love.

Here's to hoping fp16 will be more common and vram prices go down

Let's be clear, most Americans like paycheck to paycheck and can't afford losing a job or being arrested for saying antigoverment stuff