'Five Open-Source Projects AI Enthusiasts Might Want To Know About'

b-dayyy quotes Linux Security: As AI becomes more and more ingrained in our daily lives through consumer products, we can't help but be concerned that proprietary software will comprise the market. And we are not talking about a million-dollar market, but a bigger one that may reach US$118.6 billion by 2025. Many industries and end-users would thus benefit from more open-source AI projects and tools for developers' use. That would save tons of individuals and companies money to build their own AI-powered apps.

In this post, we explore five open-source AI projects or tools that are compatible with Linux and delve into the pros and cons of open-source AI and AI in general.
The list includes TensorFlow by Google's AI research team, as well as Microsoft Cognitive Toolkit. The article points out that open-source AI "is also being explored in developing hardware, specifically microprocessors that are more secure," and suggests some other possible transformative uses (including smart farming technologies "that aid in livestock and crop monitoring, irrigation, weather forecasting, and overall farm management... [H]ealthcare becomes more factual than intuitive, increases in revenue can be seen more clearly in marketing efforts, and food security becomes a reality rather than a dream.

"However, we should not discount the fact that AI can also be weaponized, empowering the wrong people. Cybersecurity systems must also be upgraded to counter AI-powered cyberattacks. And when developing AI-powered machines, it is critical to ensure that they are not vulnerable to attacks."

I was hoping for a new

[bobstreo]

Voice Recognition system that didn't depend on anything in the cloud (or off the system it's used on) to do it's job.

Or maybe some text to speech tools that actually sounded like a human, to go with the VR above.

Re:

[Actually, I do RTFA]

There have been a lot of VR systems that worked well-enough offline. Hell MS gave one away free with XP. It needed 20 minutes of training and you needed to switch voice models manually, but it worked pretty well.

Re:

[aeyec]

For AI use, "works pretty well" is worthless. And Win10's VR is worse than useless. I spent a lot of time trying to get it to work well enough to use with my AI software. It's easier for me just to type stuff in.

Re:

[Actually, I do RTFA]

Did you train it? XP's worked pretty well once trained. Meaning dictation and correction was better than typing for people who don't type well, and it worked well enough for commands that it was reliable in the field.

Like, I'm thrown. If you wanted a natural language AI, sure it wasn't up to that. If you wanted it to respond with ll the menu items and keyboard shortcuts, it was.

CNTK

[serviscope_minor]

WTF Microsoft's toolkit? That came before PyTorch?

I've never met anyone who's used it it even knows someone who uses it.

Tensorflow?

[Njovich]

Yeah I think most AI enthusiasts are going to be aware of tensorflow. Here is five technologies tech enthusiasts should be aware of: computers, mobile phones, cars, light bulbs, televisions.

Re:

[phantomfive]

I came to say the exact same thing. You don't know about Tensorflow? Where have you been the last few years?

Also notable: PyTorch is not on the list. What kind of list is it?

AI?

[DontBeAMoran]

There's no such thing as "AI" just yet. All they're toying around with is pattern recognition, there's no intelligence in that.

Re:

[Njovich]

There is no such thing as 'intelligence', hate to break this to you but humans are nothing more than deterministic pattern recognizers too.

Re:

[mrclevesque]

"There is no such thing as 'intelligence' "

https://en.wikipedia.org/wiki/... [wikipedia.org]

"humans are nothing more than deterministic pattern recognizers too."

They are rather different in how they do it.

Re:

[phantomfive]

humans are nothing more than deterministic pattern recognizers too.

Human brains are Turing Complete, whereas most neural networks are not.

Re:

[Antique Geekmeister]

While the irony in your comment is clear, nerves are not deterministic, a great deal of neural behavior is not deterministic. It's too vulnerable to random effects that occur after the original state of the system is set and all inputs controlled, due to thermal noise and quantum level effects of electrical impulses triggering or inhibiting other events in a spatially and temporally sensitive way.

Re:

[aeyec]

While the irony in your comment is clear, nerves are not deterministic, a great deal of neural behavior is not deterministic. It's too vulnerable to random effects that occur after the original state of the system is set and all inputs controlled, due to thermal noise and quantum level effects of electrical impulses triggering or inhibiting other events in a spatially and temporally sensitive way.

Well, he could have left out "deterministic", but "pattern recognizers" is true.

Re:

[Aighearach]

AI is a department at the University, where they study various classes of algorithms.

So it exists.

The real problem with AI is that they're trying to train everything now, and work in expert systems has slowed as a consequence.

Re:

[Kjella]

There's no such thing as "AI" just yet. All they're toying around with is pattern recognition, there's no intelligence in that.

Except that we do consider this a sign of intelligence. This is Pavlov's classic experiment, if you sound a dinner bell before dinner some species like dogs will take note and start anticipating food through salivation and going to the food dispenser. Other more primitive species will lack the intellectual capacity to make the connection and only seek out the food once it's served. That is of course after we've verified they have the physical capacity to make the distinction, of course if they can't hear th

Re:

[phantomfive]

Well, the pattern recognition algorithms are kind of cool, though.

Re:

[Paradise Pete]

There's no such thing as "AI" just yet.

And there never will be, judging by how the goal posts keep moving.

My AI already told me weeks ago

[nospam007]

You can't rely anymore on those pesky humans.

Mentioning a dying Tensorflow?

[Jarwulf]

but no mention of Pytorch which is becoming the defacto general purpose ml library? Sounds like just a blog randomly posting things they heard about but don't really understand.

Re:

[phantomfive]

I don't think you can really call Tensorflow "dying"

Uhh, duh.

[Gravis Zero]

"However, we should not discount the fact that AI can also be weaponized, empowering the wrong people.

Who do you think is paying for all the research? Do you honestly think they aren't already weaponizing it?

Re:

[cusco]

increases in revenue can be seen more clearly in marketing efforts

Marketers? Definitely not the people we want getting into AI . . .

Re:

[ElizabethGreene]

> Marketers? Definitely not the people we want getting into AI.

I respectfully disagree. I find value in product suggesters, the little "You might also be interested in..." or "based on your previous purchases" things in online stores. That is an actually helpful marketing AI, and one I welcome.

Darknet

[mejustme]

Another great open source one for linux missing from that list: Darknet. https://github.com/AlexeyAB/da... [github.com] Home of YOLO ("you only look once") realtime object detection.

Re:

[onceuponatime]

Darknet does indeed utterly rock. Here's the link to the original darknet as well:

https://pjreddie.com/darknet/ [pjreddie.com]

Kim

Lame

[invalid_user]

As someone who follows a ton of AI researchers on Twitter, I find this list totally out of touch with the reality on the field.

I know Slashdot is read by old timers, but this is lame.