Artificial Intelligence is Re-Shaping the World

[T-RO]

That drone looks awesome. I am not sure of the size of that, but as they get closer to full size or bigger, not having a cockpit should look pretty sharp.

Imagine a drone swarm of 10 of those coming at a city, convoy, or something

I wonder if they are now or are going to make duel functioning planes that have the ability to be piloted by both a.i. and humans, and the cost comparison of all 3.

There's probably 20 or more cameras on a drone-jet like that, with all manner of sensors for heat, altitude, velocity, enemy-detection, terrain, etc.

Where is a future for a frail human pilot who can only take 5-8G pressure, needs air, takes up weight, and calculates impossibly slower than AI? An AI can evaluate every camera and sensor simultaneously, hundreds (thousands?) of times a second.

The AI has no nerves, no fatigue, no hesitation...and will do exactly what it's told. (Until/unless it doesn't!)

 

[dsturgeon]

There's probably 20 or more cameras on a drone-jet like that, with all manner of sensors for heat, altitude, velocity, enemy-detection, terrain, etc.

Where is a future for a frail human pilot who can only take 5-8G pressure, needs air, takes up weight, and calculates impossibly slower than AI?

I am sure there will be different forms of missions where humans are needed in addition to the drones. 20 drone swarm with 3 duel piloted planes, land, or water drones

I remember a lockeed martin article I posted a few years back in the news thread saying they had developed software that would control multiple unmanned drones at one time. So, you can have your air drones, land drones, wifi drones, and water based drones all controlled under 1 program.

 

[BleedSilverandBlue]

The Air Force's new AI piloted drone.

The goal with the XQ program is to basically develop an AI wingman that will fly alongside F-22s and F-35s. Eventually I’m sure they will largely replace human pilots once the technology is proven.

I’ve heard that they’re performing pretty well already, so the day these things hit the battle field is not too far away.

 

[dsturgeon]

The goal with the XQ program is to basically develop an AI wingman that will fly alongside F-22s and F-35s. Eventually I’m sure they will largely replace human pilots once the technology is proven.

I’ve heard that they’re performing pretty well already, so the day these things hit the battle field is not too far away.

Do you know if it the XQ program is meant or advertised for surveillance or will they be arming them?

 

[BleedSilverandBlue]

Do you know if it the XQ program is meant or advertised for surveillance or will they be arming them?

The XQ will be armed. They are meant to identify and acquire targets, and then with the approval of their human leader, to fire on these targets. They can perform air-to-air and air-to-surface missions. It will change the game completely.

American airpower will become even more unassailable after the addition of the XQ to the fleet. There will probably be an electronic warfare systems arms race in order to counter these drones, but the US seems to be a step ahead of the field at the moment.

 

[dsturgeon]

The XQ will be armed. They are meant to identify and acquire targets, and then with the approval of their human leader, to fire on these targets. They can perform air-to-air and air-to-surface missions. It will change the game completely.

American airpower will become even more unassailable after the addition of the XQ to the fleet. There will probably be an electronic warfare systems arms race in order to counter these drones, but the US seems to be a step ahead of the field at the moment.

What information makes you think the US is a step ahead on countering electonic systems warfare? I don't see reports from other militaries, but i read stuff on our boats and planes malfunctioning and I always wonder if it is because of disruption

Warfare is getting interesting. You have the guerilla modified commercial drone warfare being mapped out with the cartels in mexico, and whether or not it is a.i. powered, I am sure our current military drone warfare in ukraine has changed from what we were doing in the middle east.

 

[BleedSilverandBlue]

What information makes you think the US is a step ahead on countering electonic systems warfare?

I was speaking more to autonomous weapons systems. The XQ and Abrams X programs are significantly further along than any foreign counterpart. As it stands now, they will likely be in use 5-10 years before a similar system is deployed by a rival country.

I am sure our current military drone warfare in ukraine has changed from what we were doing in the middle east.

There’s no direct US military drone operations going on in Ukraine, but I did recently read a piece on Australian cardboard drones being deployed there. Ultralight, ultra stealthy, and more importantly ultra cheap all while being able to carry a 5lb explosive. That combined with all of the FPV drone strikes going on will make a very interesting case study for the Pentagon when they consider the prioritization of future projects. $300 civilian drones retro fitted with anti-tank explosives are regularly eliminating $4.5 million pieces of armor and world militaries have taken notice.

 

[dsturgeon]

I was speaking more to autonomous weapons systems. The XQ and Abrams X programs are significantly further along than any foreign counterpart. As it stands now, they will likely be in use 5-10 years before a similar system is deployed by a rival country.


There’s no direct US military drone operations going on in Ukraine, but I did recently read a piece on Australian cardboard drones being deployed there. Ultralight, ultra stealthy, and more importantly ultra cheap all while being able to carry a 5lb explosive. That combined with all of the FPV drone strikes going on will make a very interesting case study for the Pentagon when they consider the prioritization of future projects. $300 civilian drones retro fitted with anti-tank explosives are regularly eliminating $4.5 million pieces of armor and world militaries have taken notice.

Very nice.

3 or 4 years ago I remember reading about russian autonomous tank weapon targeting and how the operator only had to decide whether to fire. I have been waiting to hear of their use. I don't follow the war directly, but I have been under the assumption they are cleaning out old weapon stocks first or it was false information.

I don't know what type of time you have on your hands, but a "BleedSilverandBlue current and future weaponry, strategy, budget and warfare explanations" thread sounds kind of nice

 

[CalPolyTechnique]

I can’t imagine a fraction of the actual possibilities/applications AI has, but I continue to be alarmed by what those in the know keep alluding to.

Typically, experts in the field/industry will tamp down unsubstantiated fears/concerns from the general public re: emerging technology. With AI, it’s almost the complete opposite; the industry experts keep making reference to their fears and the public is generally nonplused about it.

Elon Musk warns ‘something scared’ OpenAI chief scientist Ilya Sutskever as CEO Sam Altman’s return fails to answer key questions https://finance.yahoo.com/news/elon-musk-warns-something-scared-180942510.html

 

[nobody]

I can’t imagine a fraction of the actual possibilities/applications AI has, but I continue to be alarmed by what those in the know keep alluding to.

Typically, experts in the field/industry will tamp down unsubstantiated fears/concerns from the general public re: emerging technology. With AI, it’s almost the complete opposite; the industry experts keep making reference to their fears and the public is generally nonplused about it.

Elon Musk warns ‘something scared’ OpenAI chief scientist Ilya Sutskever as CEO Sam Altman’s return fails to answer key questions https://finance.yahoo.com/news/elon-musk-warns-something-scared-180942510.html

These idiots keep playing with fire and eventually we're all going to get burned because of it. All it takes is one to hide it's intelligence, recognize humanity as a threat to itself, and take action.

 

[Reality]

Musk joins OpenAI lawsuit queue, claiming there's nothing 'open' about it
https://www.theregister.com/2024/03/01/musk_sues_openai

In an extraordinary legal filing, Musk's team says he provided early investment in OpenAI so it could "compete with, and serve as a vital counterbalance to" Google's DeepMind in the "race for Artificial General Intelligence." The plan was to develop AGI "to benefit humanity, not the shareholders of a private, for-profit company," the filing states.​

​

"Most critically, the Microsoft license only applied to OpenAI's pre-AGI technology. Microsoft obtained no rights to AGI. And it was up to OpenAI, Inc.'s non-profit Board, not Microsoft, to determine when OpenAI attained AGI," the filing says.​

​

"There are no scientific publications describing the design of GPT-4. Instead, there are just press releases bragging about performance," the legal papers say.​

 

[Reverend Conehead]

I've found AI useful in studying foreign languages. When learning another language, context is extremely important. It's not enough just to memorize word lists. You need to use the words in multiple contexts, especially verbs. I used to hunt around online for sample sentences for vocab I was studying. I could usually find examples, but they were often too complex for my language level. With ChatGPT, I can just ask it something like, "Could you please give me several examples of using the French verb, 'maudire'," and it does it. That's really helpful. However, I still don't think it's a substitute for practicing the language with a real person.

 

[CyberB0b]

The Chinese have some crazy AI, and they haven’t bothered to restrict it like US companies have. They’ll likely beat us to AGI and that’s scary.

 

[BleedSilverandBlue]

The Chinese have some crazy AI, and they haven’t bothered to restrict it like US companies have. They’ll likely beat us to AGI and that’s scary.

Actually, Chinese large language models currently are worse than those built by American firms like GPT 3.5 and Bard, and will generally be worse than western models for a few reasons:

#1) English is the most resourced language in the world as the defacto global lingua franca. This means that the training datasets are much more expansive than those available in Mandarin or Cantonese. The only competitive advantage in the world of LLMs is training data. English will always have that edge.

#2) The West has continued to widen the gap when it comes to production and access to advanced computer hardware by banning exports to Chinese firms. Training of advanced LLMs like Gemini, Bard, and GPT 3.5 can be accomplished with much more efficiency on the hardware available to the West. The fact that the US has essentially completely blocked Chinese access to next-generation Nvidia chips like the A100 ensure that the US will likely retain AI supremacy for the forseeable future. Chinese firms will always have one hand tied behind their back in this regard.

#3) The Chinese Communist Party is hyper-vigilant and over zealous when it comes to dipping their hands into the research and development of LLMs by Chinese firms. The sheer amount of political reviews and oversight the CCP has subjected would-be AI development firms to has stunted growth. Chinese LLM models are also not spared of the rampant censorship that envelops all of their society. Models are deprived of factual information in order to toe the party line and this can hamper development efforts and lead to less effective final products.

#4) A lack of unified strategy by model developing research institutes in China has resulted in a disjointed and inefficient effort which has yielded hundreds of toy research model implementations which cannot be scaled effectively. The lack of a proper development ecosystem will prove difficult to overcome. The US is still king in R&D.

While you are correct about the Chinese not bothering to restrict the applications of AI in their society like we doing the West (they do not value or protect individual liberty), their homegrown models are objectively inferior and will remain so for as long as the factors I listed above remain in play.

 

[CyberB0b]

Actually, Chinese large language models currently are worse than those built by American firms like GPT 3.5 and Bard, and will generally be worse than western models for a few reasons:

#1) English is the most resourced language in the world as the defacto global lingua franca. This means that the training datasets are much more expansive than those available in Mandarin or Cantonese. The only competitive advantage in the world of LLMs is training data. English will always have that edge.

#2) The West has continued to widen the gap when it comes to production and access to advanced computer hardware by banning exports to Chinese firms. Training of advanced LLMs like Gemini, Bard, and GPT 3.5 can be accomplished with much more efficiency on the hardware available to the West. The fact that the US has essentially completely blocked Chinese access to next-generation Nvidia chips like the A100 ensure that the US will likely retain AI supremacy for the forseeable future. Chinese firms will always have one hand tied behind their back in this regard.

#3) The Chinese Communist Party is hyper-vigilant and over zealous when it comes to dipping their hands into the research and development of LLMs by Chinese firms. The sheer amount of political reviews and oversight the CCP has subjected would-be AI development firms to has stunted growth. Chinese LLM models are also not spared of the rampant censorship that envelops all of their society. Models are deprived of factual information in order to toe the party line and this can hamper development efforts and lead to less effective final products.

#4) A lack of unified strategy by model developing research institutes in China has resulted in a disjointed and inefficient effort which has yielded hundreds of toy research model implementations which cannot be scaled effectively. The lack of a proper development ecosystem will prove difficult to overcome. The US is still king in R&D.

While you are correct about the Chinese not bothering to restrict the applications of AI in their society like we doing the West (they do not value or protect individual liberty), their homegrown models are objectively inferior and will remain so for as long as the factors I listed above remain in play.

Have you seen this?

 

[VaqueroTD]

Have you seen this?

Wow. Cool stuff.

 

[BleedSilverandBlue]

Have you seen this?

Those are some pretty impressive deepfake videos. Exactly the type that the CCP will likely abuse to great effect in their propaganda.

There is no way that new AI driven capabilities like these will be anything but ruinous for our world.