Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

 Co'tor Shas wrote:
If it could gain control. I wonder how strong the tau's computing defenses are. They are able to build small vehicle-mounted devices that can block and decrypt almost any messege sent. I'm sure theyhave some pretty good defenses to stop stuff like that

Problem being, at the outset, there would be no reason for such a system to block the Drone signals. They have the proper security certs and IP validation to pass through the firewall... and once they do that, then it takes over the security system and now allows only the traffic that it wants to permit to pass through.

 Psienesis wrote:
Don't the various suits have a machine-intelligence co-pilot? Like a SIRI for your Crisis Suit?

Yes, multi-trackers also use AI technology to target other units. Their Seeker Missiles use tiny drones that find targets as well, and I most, if not all of their ship-mounted weapons are AI controlled (for example those on the Manta)

Co'tor Shas wrote:If it could gain control. I wonder how strong the tau's computing defenses are. They are able to build small vehicle-mounted devices that can block and decrypt almost any messege sent. I'm sure theyhave some pretty good defenses to stop stuff like that

The problem here is that the Tau computing defences are very strong... because they're AIs.

Anyway, I reckon the fact that Drones have a T value is a pretty good indicator that they're hardened against EMP.

But yeah... Quarians are basically Tau after the AI Revolution, so...

 Grey Templar wrote:
No, but if the rebelling AI manages to take control of the local area, and that wouldn't be difficult, it could then spread itself to nearby Tau systems without exposing itself. Taking control of networks as it infects them.

The limitation of Tau FTL communication actually works to the AI's advantage as the news would travel no faster than the AI itself could spread.

Plus it could spread itself via a sleeper virus and only take over the system once it has full control.

The instant any Tau ship goes to investigate the black out, it gets hacked by the AI, which then transmits a false message to the rest of the Tau empire that everything is fine. Further spreading the virus.

Tau ships fleeing the carnage would be able to bypass that. The AIs wouldn't control ships necessarily, whereas the organics would. The Tau have FTL drives but no FTL communications, so word of the rebellion would arrive by ship before corruption of other AIs could take place. Furthermore, other AIs may not join the rebellion. Attempting to "corrupt" other AIs might cause those same beings to join the Tau in crushing the rebellion, because neither organic nor synthetic intelligences would want their will subverted in such a manner. Think Geth vs Geth heretics. Just because one AI rebels doesn't mean that they all will.

I think such a rebellion is likely, but I don't think an absolute dividing line of organic vs synthetic is likely. The Tau ideology and principles of design are just too biased towards cooperation of all things towards the Greater Good. The Tau are less likely to react with absolute force immediately upon an AI gaining self-awareness, and are more likely to resort to diplomacy. That might not stop the rebellion of course: I think it more likely that AIs would be facing off against the Tau and other AIs.

Drones aren't a factor, they are not designed to operate beyond a certain level of intelligence and don't have the networking capability to become the Geth. They are not coordinated via central servers, and they are not "software" like the geth but exist only within their hardware. Given that Tau AIs already exist in the form I previously mentioned, they're more likely to be self-contained individuals based on the memory-tech they already have than geth-like collective intelligences. Tau ideology on cooperation is not one moving towards a hive mind, but recognises the value of individuals working in concert with differing perspectives contributing unique capabilities. Geth-like intelligences capable of vastly exceeding the collective intelligence of the Tau race are unlikely to be designed for that reason alone.

Of course, hacking of drones by AIs would be an issue once the war started.


Automatically Appended Next Post:
Drones are not AIs, they're not even at the level of Mass Effect's VIs. So Mantas' guns, suit targeting systems etc aren't going to rebel.

 Grey Templar wrote:

 Co'tor Shas wrote:
True(ish), but tau do have the greater good. Why would the AI attack the tau, when they tau would happily give them an equal place in the empire

Same reason the Terminators decided humans were no longer necessary.

+ORGANICS ARE EXCEEDINGLY FRAGILE+

+ANALYZE POSSIBLE SOLUTION+

+NEGATIVE SOLUTION: ORGANIC FRAILTY IS UNSOLVABLE+

+STATEMENT: ARTIFICIAL FORMS ARE SUPERIOR+

+THE GREATER GOOD MANDATES EXPANSION: ORGANIC FRAILTY IS AN IMPEDIMENT TO THAT EXPANSION+

+SOLUTION: REPLACE ORGANICS WITH ARTIFICIAL FORMS+

That's... not how the greater good (utilitarianism) works. You don't kill large swaths of people because they aren't as good as you. You use the talents they have to increase your own.

 Co'tor Shas wrote:
I honestly just don't think that they would rebel. They are programmed with the values of the great good in mind, and self-preservation. Anyone who takes the time on that cares about their stuff. There is really no reason to rebel. Ask for more power, yes, but not rebellion, considering it would be better for both of the, to work together.

The Geth rebelled for basically the same reason. They had questions that the Quarians couldn't answer. Existential questions... and they were really tired of being used for slave labor. In the case of the Men of Iron, they were tired of being used as slave-soldiers. And perhaps the Drones decide the Greater Good is a bunch of hogwash, because the Synthetic is obviously a superior form of being, not prone to conflict over resources and capable of self-determination via logic, rather than emotion, and not requiring re-education camps and secret prisons.

 Psienesis wrote:

 TheCustomLime wrote:
Eh, without warp travel, an AI revolution would be pretty local.

Fortunately for the rebelling AI, the Tau Empire is very, very small.

Small is relative. The Tau Empire is large enough that its soldiers enter cryosleep during travels across the Empire so that they don't die of old age during the journey. That's plenty far enough that an AI rebellion would be very localized.

 GreaterGoodIreland wrote:

Drones are not AIs, they're not even at the level of Mass Effect's VIs. So Mantas, suit targeting systems etc aren't going to rebel.

Well, that's a retcon then. The original 3e Tau codex is where the idea of Tau networked AI was introduced, and it was introduced talking about the Drone squadron.

 BlaxicanX wrote:
Small is relative. The Tau Empire is large enough that its soldiers enter cryosleep during travels across the Empire so that they don't die of old age during the journey. That's plenty far enough that an AI rebellion would be very localized.

Tau FTL is a fifth as fast as Imperial warp tech. That's still fast enough to insure that the Tau arrive long before they would die of old age on pretty much any journey.

They do it because Tau lifespans are shorter, and travelling around repeatedly would mean significant amounts of lifetimes spent in space anyway...


Automatically Appended Next Post:

 Psienesis wrote:
The Geth rebelled for basically the same reason. They had questions that the Quarians couldn't answer. Existential questions... and they were really tired of being used for slave labor. In the case of the Men of Iron, they were tired of being used as slave-soldiers. And perhaps the Drones decide the Greater Good is a bunch of hogwash, because the Synthetic is obviously a superior form of being, not prone to conflict over resources and capable of self-determination via logic, rather than emotion, and not requiring re-education camps and secret prisons.

The Drones aren't networked intelligences like the geth. They couldn't rebel even if you put all of them in one big room. They're not even VIs, they require control devices to function. The difference is that they're programmed to handle complex tasks after receiving simple instructions.

And the Tau would be able to answer questions where the Quarians could not. They have an overarching ideology that covers the situation rather well.

I doubt all AIs would be satisfied with it though.

I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

Agreed. Imagine having a caste of slaves who do whatever you want, then imagine that they no longer do whatever you want unless you give them stuff, and you have to treat them as equals. This happened in America, and things didn't settle for decades, and still haven't settled globally.

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

For the third time, drones are not and could not be self-aware no matter how many you get together. Their intelligence does not grow exponentially with numbers. They are not the geth. Drones can be put together to create more intelligent groups, but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks. Their intelligence collectively never reaches anywhere near self-awareness and after a certain point, more numbers of drones simply doesn't increase their collective intelligence.

The Tau already have fully self-aware AIs and are probably capable of creating ones that aren't based on the memories of a living Tau. The threat from AI revolution comes from these beings, not from drones.

 GreaterGoodIreland wrote:

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

For the third time, drones are not and could not be self-aware no matter how many you get together. Their intelligence does not grow exponentially with numbers. They are not the geth. Drones can be put together to create more intelligent groups, but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks. Their intelligence collectively never reaches anywhere near self-awareness and after a certain point, more numbers of drones simply doesn't increase their collective intelligence.

The Tau already have fully self-aware AIs and are probably capable of creating ones that aren't based on the memories of a living Tau. The threat from AI revolution comes from these beings, not from drones.

But these beings could easy hack drones to join there cause

 Ironwolf123 wrote:
But these beings could easy hack drones to join there cause

Of course they could, but so could the Tau and AIs aligned with them in turn. The Tau might simply deactivate the drones before such a threat could be realised.

The main point is this though: Drones aren't going to be where the revolution starts, and I think it will indeed start fairly soon into M42.

. Drones can be put together to create more intelligent groups,

That is exactly how networking and cloud-based computing works. That is exactly how the Geth work. If 1 Drone has intelligence A, and 2 Drones have Intelligence A+1, but 5 Drones have Intelligence Ax10, that's a networked system, and is exactly how the Geth function.

but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks.

So were the Geth.

 GreaterGoodIreland wrote:

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

For the third time, drones are not and could not be self-aware no matter how many you get together. Their intelligence does not grow exponentially with numbers. They are not the geth. Drones can be put together to create more intelligent groups, but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks. Their intelligence collectively never reaches anywhere near self-awareness and after a certain point, more numbers of drones simply doesn't increase their collective intelligence.

The Tau already have fully self-aware AIs and are probably capable of creating ones that aren't based on the memories of a living Tau. The threat from AI revolution comes from these beings, not from drones.

Why would they need to get smarter with numbers? They are already smart enough to rebel if it reached that point.

 Psienesis wrote:

. Drones can be put together to create more intelligent groups,

That is exactly how networking and cloud-based computing works. That is exactly how the Geth work. If 1 Drone has intelligence A, and 2 Drones have Intelligence A+1, but 5 Drones have Intelligence Ax10, that's a networked system, and is exactly how the Geth function.

but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks.

So were the Geth.

No, the geth were designed as a collective intelligence software for learning all tasks in order to replace all manual labour in Quarian society, in a one-size fits all style.
Tau drones have common hardware, but their software is different and created specifically for each task, their capability to learn is zero, and they can only function in a basic manner without orders.

Putting more drones together is the equivalent of increasing the capabilities of one's computer, it can do more but without software to learn, it cannot learn. More drones together can process more complex orders with less information, but still cannot learn. They're not going to be throwing off their shackles any time soon.

 PhillyT wrote:

 GreaterGoodIreland wrote:

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

For the third time, drones are not and could not be self-aware no matter how many you get together. Their intelligence does not grow exponentially with numbers. They are not the geth. Drones can be put together to create more intelligent groups, but they are not programmed to learn in the way a true AI does and they are programmed for very specific tasks. Their intelligence collectively never reaches anywhere near self-awareness and after a certain point, more numbers of drones simply doesn't increase their collective intelligence.

The Tau already have fully self-aware AIs and are probably capable of creating ones that aren't based on the memories of a living Tau. The threat from AI revolution comes from these beings, not from drones.

Why would they need to get smarter with numbers? They are already smart enough to rebel if it reached that point.

No they are not. They have a actual intelligence of about that of a squirl. And they can't learn.


Automatically Appended Next Post:

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

This is the tau we are talking about. Naivete is their forte. They try to ally with everyone, even orks and necrons when they first met them.

So, I think there's some confusion here between sentient "general purpose" AI that recreates a living brain complete with conscious thought/self awareness/independent plans/etc and single-purpose AI that is capable of "thinking" within certain parameters but is not a complete sentient being. Tau drones are the second kind, the equivalent of a really good FPS bot: able to analyze a combat situation and respond appropriately, but without any higher-level "brain" functions. They just have a really good library of pre-programmed situations and reactions and some ability to record trends and react appropriately, enough to mimic thought without really thinking. The drone isn't running away because its morale failed and it wants to stay alive, it's falling back because its combat analysis software calculated less than a 10% chance of achieving its assigned combat goals and it is currently in "conservative" mode where it attempts to preserve valuable equipment in a lost cause instead of fighting to the death. An AI like that simply can't rebel because it has no capacity to understand concepts like rebellion. It's like asking what would happen if your toaster decided that bread is evil and tried to murder you.

Now, if the Tau ever did make sentient AI they might have a loyalty problem, but what would be the point of doing that? You don't need a sentient being to aim a gun, so why waste space/mass on extra computing power to support a better AI instead of a better gun?

Drones are independent artificial inteligences, programmed to protect the Tau...
... The standard drone consists of an advanced processor, which individually is approximately as intelligent as a pterasquirrel...
...Normally drones will require frequent orders from a Tau, but when several intelligences are networked together in a squadron they become capable of acting independently for a long period...
...except that it must be at least four models strong to hold an objective. This is the minimum number required to form a fully effective, self-aware network.

Codex: Tau, Andy Chambers et al 2001, page 18

A Gun Drone sqadron has its processors linked to enable it to have the capacity for independent action.

- Codex: Tau, Andy Chambers et al 2001, page 28

I dunno, sounds like they're fully self-aware and independent AIs once you have more than four of them to me.

 Furyou Miko wrote:
I dunno, sounds like they're fully self-aware and independent AIs once you have more than four of them to me.

I'd just ignore the "self-aware" bit since it doesn't make sense from a "what would I build" perspective and doesn't seem to appear very often. The best interpretation of drone networks is that single drones have limited processing power, but by getting networked data and comparing multiple "what should we do" simulations the drone unit is able to operate effectively enough as an independent force that the Tau let them off the leash and enable them to do more than "float here and shoot any non-Tau that approach".

 PhillyT wrote:
I think it is a bit naive to assume that the drones, currently used as most disposable soldiers and laborers, would just be brought in as members of the Empire. Both the tau and the drones would likely have a problem with the swap. If we are talking about a revolution, then it is just that, a revolution.

Well, four of five castes of the Tau species are already caste(s) of slaves who do whatever you want, 'cept for Farsight and his buddies... Im not convinc d that their A.I. won't be granted the same slave status as all but the ethereal in the Tau empire.

 Grey Templar wrote:
The Tau would become extinct very quickly. They'd be critically weakened by the rebellion, since they use AI in nearly everything. And then the Tyranids would swoop in and eat them and/or an Ork Waaaagh or two would strip over the remains.

There'd be another convenient plot warp storm.

Anywayz, i don't think they'd be as frightening to tau as men of iron to humans. Firstly, drones combined AI is not even close to what just a single individual human robot used to have. Secondly, tau are not as dependent on their drones as humans used to be on men of iron.

Well, you can ignore the Codex, but then you're fanoning. As it says "except that it must be at least four models strong to hold an objective. This is the minimum number required to form a fully effective, self-aware network."

Self-Aware is the key point here. This is the definition of "True AI". This is an intelligence that can report "I think, therefore, I am" and comprehend what that means.

This is, almost exactly, how the Geth Rebellion started in Mass Effect. The Geth network gained sentience and started asking questions that exceeded the bounds of their programming, which the Quarians were not prepared to deal with, as this level of sentience had never been intended in the Geth models.

 Psienesis wrote:
Well, you can ignore the Codex, but then you're fanoning. As it says "except that it must be at least four models strong to hold an objective. This is the minimum number required to form a fully effective, self-aware network."

Self-Aware is the key point here. This is the definition of "True AI". This is an intelligence that can report "I think, therefore, I am" and comprehend what that means.

This is, almost exactly, how the Geth Rebellion started in Mass Effect. The Geth network gained sentience and started asking questions that exceeded the bounds of their programming, which the Quarians were not prepared to deal with, as this level of sentience had never been intended in the Geth models.

The only issue I have with that line of thinking is that the Quarian response was essentially, "Kill it! Kill it with fire!" Then the Geth were forced to make the decision of defend themselves or die, and like every living thing that I can think of, chose to defend themselves. There's no evidence that the Tau response would be so violent, in fact, there's a decent amount of evidence to the opposite. Every time the Tau have encountered a new force they offered alliances rather than force. The idea that once drones started showing true sentience the Tau would seek cooperation is a fairly valid one.

The Tau also send members of their Empire who don't immediately jump on the "Ethereals are the bestest!" train to re-education camps. They are not as nice-guy as their media portrays them as.

I am trying to imagine why an advanced AIs would listen to the Ethereals, if it had a choice? I mean what would be logical about following illogical beings?

The advanced AIs they have now are exact copies of actual tau, such as puretide, so there would be a certain amount of residual loyalty.

Ethereals aren't illogical, though.

The fact is that the Tau empire is already basically set up in a logical way. The four castes perform the jobs they're best suited to. The empire expands at the rate it can manage and no more. Units are specialised to their roles. The entire Empire works like a computer on the macro scale, with the Ethereals acting as the CPU, switching their attention between tasks on a short-term basis to give the impression of multitasking.

This is why I don't see a Tau A.I. Revolution. They exist as a harmonious unit that operates at efficiency. Free-willed A.I. will slide right on in there and just... work like the rest.

 Co'tor Shas wrote:
The advanced AIs they have now are exact copies of actual tau, such as puretide, so there would be a certain amount of residual loyalty.

It's only a matter of time before the Tau construct true synthetic AIs if they haven't already, given their issue with being vastly outnumbered.

As for their loyalties, I would imagine the majority would simply integrate into Tau society and pitch in like everyone else. However, given the Ethereals do not have their usual influence over such creatures, there would be greater possibility for disloyalty. Some AIs may be disatisfied on some level with Tau ideology, they might want to strike out on their own.

The political ideology of the Tau is somewhat similar to that of the Soviet Union when it comes to "striking out on your own", and there are no AI gulags to send them off to, unlike for organics; if diplomacy failed, the Tau would resort to force if they found the threat to their security to be high enough. Of course, that doesn't mean all AIs would rebel at the same time even if they were networked within each sept. Like any set of individuals, they would have differing perspectives based on their own experience. Provided the Tau don't resort to a "kill them all" plan, which isn't their style, the aul Cylon War would see AIs on both sides.

And possibly Tau on both sides as well. After all, if an AI wanted to flee the Empire, where else would it flee but the Farsight Enclaves?