Xenon needs a upgrade.

[jlehtone]

But going back to t he mechanic all this means that xenons mechanically are far weaker than they should be. And that applies to player-capturable ships.

The F, B, H are old designs and them retaining space for human pilot is acceptable, but I see no rationale to justify the PE and SE to be capturable.

IMHO, Xenon having some better tech could be ok, but player access to it should not be an option.

[vvvvvvvv]

The F, B, H are old designs and them retaining space for human pilot is acceptable, but I see no rationale to justify the PE and SE to be capturable.

IMHO, Xenon having some better tech could be ok, but player access to it should not be an option.

Other ships were capturable in rebirth, and I believe most of them couldbe captured in x3. You could also build them in x3 after reverse engineering in HQ. Practically the only real reason you can't cap them in x4 is because they have no cockpit.

F/B/H should have ability to retrofit better weapons, and there should be better equipment. At least Boso Ta or someone else should be able to produce it. An interesting detail: You can glitch encyclopedia to see what would happen if you installed different weapon on B, but can't install it in practice. So for some ships you end up with a single possible loadout, which is quite sad.

I see no point in restricting player access, as this is a sandbox, and one of the ideas was "if you can see it, you can fly it".

[HelgeK]

Human-level sentience isn't possible because they don't have the same sort of inputs.

I can only speak for the "novelverse", which may have some differences to actual gameplay. Reason being, what works perfectly in-game, might appear slightly silly in a novel. I'm not a gameplay expert, so I can't speak directly to that.

That being said, my take on machine sentience has always been that sentience is substrate-agnostic. Meaning, wetware or silicon, both can plausibly implement sentience. I also hold that the minimum requirements for sentience are as of yet unknown. Are all mammals sentient or only h. sapiens? All vertebrates? What about insects? Bacteria? Viruses?

I also think that there are most likely many different ways to implement sentience, and that the subjective experience (qualia) of such entities will differ greatly from that of h. sapiens. And lastly, sentience may not even be a prerequisite for a highly intelligent agent capable of planning and executing goal-oriented actions.

The original VN Terraformers were capable of finding novel solutions to problems related to their task. A faulty update disabled the mechanism that prevented their goals from drifting (back when I wrote about that some 20 years ago, "goal drift" was being discussed in AI circles. These days, I think it's called "misalignment" or "instrumental convergence"). Consequently, the machines were able to evolve - very rapidly compared to biological entities, since machines can implement changes almost immediately, whereas biological entities cannot.

What does that mean in terms of Xenon? It means that there are a number of branches, some of which are sentient and sapient, and some of which are not sentient, but still highly intelligent. And some are tool-level without either sentience or highter intelligence.

Not sure if that helps at all, but that's a short summary of what went into my ideas for the Terraformer/Xenon.

Cheers,
Helge aka Höllge


PS: Before someone asks: LLMs are neither sentient nor sapient nor AGI

[chew-ie]

Thanks for your thougths Helge, much appreciated

PS: Before someone asks: LLMs are neither sentient nor sapient nor AGI

Nobody would ask such a thing. Hopefully.

[moako]

It's a long time ago that I read the novel "Nopileos" and I didn't start with "Farnhams legend". It would be nice to watch X-Universe as a serie. Similiar made as Expands.

I agree mostly with your reply from a science point but of course you as writer of novels have more freedoms.

Human-level sentience isn't possible because they don't have the same sort of inputs.

And lastly, sentience may not even be a prerequisite for a highly intelligent agent capable of planning and executing goal-oriented actions.

OK. The question here is; what means highly intelligent? I think that you don't need sentience or intelligence to be capable of planning and executing goal-oriented actions.
Nowadays AI are already capable to be a composer or drawer.

A faulty update disabled the mechanism that prevented their goals from drifting (back when I wrote about that some 20 years ago, "goal drift" was being discussed in AI circles. These days, I think it's called "misalignment" or "instrumental convergence"). Consequently, the machines were able to evolve - very rapidly compared to biological entities, since machines can implement changes almost immediately, whereas biological entities cannot.

There it is the explanation. I see. In principle yes but there are a lot of unanswered questions if it could withstand a thoroughly science comparison. As instance:
- Had the terraformer or later on(after the accident with the software upload) the machines access to the whole knowledge of the human races?
- What is about science devices? Had terraformer or xenon acces to the knowledge how to build such devices and how to use it?
- etc.

My point is that information alone is by far not enough. It could be that an elefant or dolphin has a higher intelligent or sentience than humans.
What humans makes special are among other our hands. Without it we couldn't have built anything and without language and paper we couldn't exchange our thought and knowledge and it wouldn't be possible to preserve our knowledge, etc. to the future.

If you speak about biological entities and implement changes. What is exactly the meaning? Do you mean implement changes within the DNA? But in principle this is possible.
For product changes it is the same. In principle you could change products each day. The reason we don't do that isn't because we couldn't do it.

[adeine]

Thanks for your thougths Helge, much appreciated

PS: Before someone asks: LLMs are neither sentient nor sapient nor AGI

Nobody would ask such a thing. Hopefully.

The fact that LLMs work at all has pretty humbling implications for biological 'sentience/sapience' though, especially essentialist takes.

[vvvvvvvv]

Thanks for your thougths Helge, much appreciated

PS: Before someone asks: LLMs are neither sentient nor sapient nor AGI

Nobody would ask such a thing. Hopefully.

The fact that LLMs work at all has pretty humbling implications for biological 'sentience/sapience' though, especially essentialist takes.

Try asking a LLM this question. See what happens. Use complex LLM, and with "thinking" enabled if available.

Consider a hypothetical. Let's say there's a robot, acting as a human.

Except... the robot is connected to a vast database which describes what it is supposed to do and how it is supposed to act in all possible situations. Meaning there isn't a situation that would cause machine to glitch.

Question.

How is this different from sentience. For an outside observer this robot is indistinguishable from a sentient entity.

And following this question...

How would I know that I am sentient or sapient. "Cogito ergo sum" only indicates that I exist. But what if self is an illusion. A temporary shortcut employed by a non-sentient process? So how would we know, that our perception of self is real and not basically a convoluted abstraction emulated on unaware hardware?

[moako]

Again, also the machines have weaknesses. Doesn't matter which angle you take in this case.

Computing power equivalent of a human brain is estimated to be an exaflop and machines with such capacity exist right now. Though they're for now large halls full of computers.

Fundamentally, however, the idea is that anything that can be digitized, can be emulated and then analyzed and optimized. Our consciousness, for the record, is evolutionary recent addition, barely a million years old. Compare it to species which existed for more. So even if we propose that human brain is unique, then replicating the unique thought process of a human is a matter or building a connectome emulator. And speaking of which... last time I tried to estimate how large a semiconductor equivalent of human brain could end up as, I arrived at an estimation of one cubic centimeter cube. It is possible that I made an error. But that ultimately means that "the machine is superior".

Intelligence, sentience and awareness of the self. Why should there be any connection to the capacity of a computer
I can't imagine that those has anything to do with computing power.

Analog signals are more precise as digital but it is easier to handle digitized data because it is simply 0 and 1. And for the computer you need 0 and 1 otherwise transistors, etc. aren't able to handle a voltage which isn't on the level of 1 or 0.
But it seems that we will have in the future more as only 1 and 0. Our scientist think they will be able to create a lot of more number between 1 and 0 with a quantum computer. Such a quantum computer is some way close to analog technology because analog signals have a much higher variety. Also it is important to remember that also analog data can be emulated, analyzed and optimized.

Without cloning facilities for a human it takes decades to grow more scientists. Self-replicating robots can build more thinking ships. Those can network, and then can bruteforce problems, machine cannot solve creatively, and that involves algorithm of creativity and human imagination. Because those too, are, fundamentally, algorithms, albeit not simple ones.

The quantity of scientists isn't that important. What we need is a good documentation, a lot of different science devices, fast computers, innovations, ideas, etc. It's about quality and not quantity.
The point about programming algorithm of creativity and human imagination can be done only as far as it focus on existing music-, picture-, etc.- samples. But that has nothing to do with real creativity and imagination.
A computer doesn't have any inputs as human do. With inputs I mean feeling, seeing, eating, enjoy, hearing, etc. All the inputs creates the creativity and imagination together with the brain(algorithm).

For example, realistically speaking upon flying into Atiya misfortunete, you should be greeted by a small strike group of a million xenon ships. In the game it is simply not going to happening, due to engine limits.

I absolutely agree. At least there could be a bit more Xenon and a bit more activity. ANT need to be stronger attacked.

The limits of machine is reproduction mechanism begin complex and requiring factories. But xenon has an answer for that. See section C-23 on assembler drones.

This section C-xx isn't really an answer to the issue and by the way assembler roboter are already in use. Of course roboter are not drones but I think that is close enough .

[vvvvvvvv]

...

> Why should there be any connection to the capacity of a computer

Because you cannot reach sapience/sentience with just one neuron. There's minimum processing power and neural network depth as prerequisite. Neural network or its equivalent.

More computing power --> bigger and more powerful brain.

> Analog signals are more precise as digital

It is opposite, analog is fuzzy logic, and definition of imprecision.

> because it is simply 0 and 1

Those 0 and 1 can represent any analog signal, and computer does not have to use binary logic.

> The quantity of scientists isn't that important.

It is, extremely even. Scientists cooperate and exchange data, and larger number of scientists increases your overall progress speed. Look at history, and note how with growth of populace there were more and more simultaneous discoveries. Meanwhile in ancient world there were single people inventing things. Quantity matters, if you have only one einstein you won't be getting anywhere any time soon. Because while einstein works on his theories an army of other scientists will be working on everything else.

> way assembler roboter are already in use.

Assembler robots matter because they reduce factory requirements. They are not used in the game, construction drones are not assembler robots described in the encyclopedia.

Xenon are von neumann probes, that means they've already mitigate greatest vulnerability of machinery where it usually needs a humongous construction complex, possibly planetary sized, to reproduce. You can also see their economy in encyclopedia. They build out of raw materials. It is extremely efficient.

[jlehtone]

How does this philosophical debate make the Xenon "better" in the game?

[moako]

> Why should there be any connection to the capacity of a computer
Because you cannot reach sapience/sentience with just one neuron. There's minimum processing power and neural network depth as prerequisite. Neural network or its equivalent.
More computing power --> bigger and more powerful brain.

You didn't answer correctly to my post before. You have simply ignorered the context of my post.
I didn't say and I didn't mean that you can reach sapience, sentience, intelligence and awareness of the self only with just one neuron.
And again, computing power has nothing to to with sapience, sentience, intelligence and awareness of the self.
Only a brain isn't worth a penny. That would be the same as a computer without energy and data.

> Analog signals are more precise as digital
It is opposite, analog is fuzzy logic, and definition of imprecision.

Due to which reason should it be imprecise
If you say it is imprecise please give arguments.

> because it is simply 0 and 1
Those 0 and 1 can represent any analog signal, and computer does not have to use binary logic.

Yes, it can represent any analog signal. That is correct. I didn't say anything different so why are you saying that
You are saying that computer doesn't have to use binary logic? You mean in theory? What exactly are you trying to say

> The quantity of scientists isn't that important.
It is, extremely even. Scientists cooperate and exchange data, and larger number of scientists increases your overall progress speed. Look at history, and note how with growth of populace there were more and more simultaneous discoveries. Meanwhile in ancient world there were single people inventing things. Quantity matters, if you have only one einstein you won't be getting anywhere any time soon. Because while einstein works on his theories an army of other scientists will be working on everything else.

Also in this case. You have simply ignorered the context of my post.
I simply said that the quantity of scientists isn't that important in context to your post before.
Of course it is important to cooperate and exchange data and a lot of other things which I also mentioned before.
Also I didn't say that quantity isn't important. I said it is about quality and not quantity in the whole context of my post before.

> way assembler roboter are already in use.
Assembler robots matter because they reduce factory requirements. They are not used in the game, construction drones are not assembler robots described in the encyclopedia.

Yes. I didn't say anything different.

Xenon are von neumann probes, that means they've already mitigate greatest vulnerability of machinery where it usually needs a humongous construction complex, possibly planetary sized, to reproduce. You can also see their economy in encyclopedia. They build out of raw materials. It is extremely efficient.

First I have to read the novel but so far it seems that I agree.

[vvvvvvvv]

And again, computing power has nothing to to with sapience, sentience, intelligence and awareness of the self.

It does, because lack of computing power limits ability to have all those things. It is a prerequisite.

Due to which reason should it be imprecise

I could flip it around. Is there a reason it should be precise? You claimed that first. "Analog is more precise than digital".

As for practice see what happens to analog signal during transmission and storage. It is distorted by environment condition. Which is why we use digital. When the only valid stats are on and off, it is much easier to deal with noise or some component slightly changing its characteristic due to temperature fluctuation.

Do you remember VHS and analog TV ghosting? That's analog in practice.

[RomanKost]

Xenon need a BIG power bust. As of today, thay rep of "a big and fearsome threat" is just laughable. They just die out as soon as a mid-game pfase of the game starts, or even before that.

Maybe they need a more powerful ships? Maybe redused construction costs? IDK

All i can say - they need an upgade and they need it badly.

[chew-ie]

Xenon need a BIG power bust. As of today, thay rep of "a big and fearsome threat" is just laughable. They just die out as soon as a mid-game pfase of the game starts, or even before that.

Agreed - as they slaughtered millions and entire fleets there has to be an endless stream mechanic out of some gates - and/or one-way accelerators like we have it in the Yaki plot. Heck - even a random gate malfunction themed story could spit out Xenon reinforcements. And / or adding the good old Xenon J incursions.

Also, in the long run there need to be new ships / a different weapon balance [for the Xenon]. Right now they don't stand a chance. Neither their stations, nor their ships.

[Ragnos28]

At this point, I do my fleet combat in X3TC, and "login" in X4 for things like ships/stations interiors, admiring the view from observation decks and so on, because in combat...
I did a 8.0 fleet combat test video, to confirm that dangerous close proximity passes from destroyers when sieging stations and carefree lingering behavior of torpedo using M ships vs capitals are alive and well (spoiler alert, they indeed are). Those tests consisted on taking out 1 I + 2 Ks and a defence station...seem that I might have to load a save prior to that, because I don't think the xenon will be able to recover those losses, with teladi filling any gap I create in xenon sectors (note: the tests were taking place in Thuruk's Demise III).

So, until the devs find a solution for "renewable" xenon, my X4 sessions will be reserved for traveling and sightseeing.

[Damian Swift]

I hate to say it (as I'm usually a purist hoping for a ventures return) but recently had to use mods just to stop the Xenon becoming extinct by mid game.

I first started with the ' Xenon jobs lite' mod which was made for a vanilla game experience to help the Xenon, but sadly even this is not enough to save them with little or no player involvement.

So I opted for the full 'Xenon Jobs' mod which gives the Xenon at least a little of the threat they used to have... They still get beaten back in the later stages of the game but at least the universe actually feels alive.

[Killrathi]

I checked how the xenons behave after the 8th update.
Don't take into mind OOS bugs/features with turrets, only logic.
About month of game with SUV, all faction warfs and ports are blocked via docks whenever possible, all plots done exept Borons.
Logic in the Southwest remains broken: free unprotected systems that the Xenons don't even try to claim.
Current map:


That is, in the last three major updates, the Xenons have been made so stupid that even with the player's help and kicks in the right direction, they move with difficulty.
This is crucial for long-term play to avoid universe overcrowding (for example, when Splits build 30-50 stations per sector). The save file grows to over a gigabyte, and at a certain point, the save simply stops loading. Not to mention the lag during gameplay with factions-owned universe.
At the same time, the Xenons typically build no more than four stations per sector: two defense stations and two powerplants. Therefore, if you give the Xenons the entire universe, leaving each faction with two or three systems, you can play for a long time without lag or the fear that saves will not loading. And the saves themselves will be about 500-600 megabytes.

PS: And one more thing. The work of S/SE ships does not play a big role. To check, you can start a custom game with XEN reputation +30 and begin too feed xenons warfs and ports with energycells, ore and silicon. Nothing will change. Xenons at center and NE will start high expancion, at SE, N and NW with madium expansion and xenon at SW will continue to be dull and remain stuck on the handbrake. Warfs and ports full of resources and absent adequate tryings to claim free free neighboring sectors.
And if you kill xenons warf and port at SW they will build new anywhere except SW sectors. So the game scripts/logic prevent the xenons progress

PPS: I just thought: there are more than 24 factions in the universe, 17 of which have built ships hostile to the Xenons. The Xenons can have no more than 7 warfs and 7 ports. We could add an increase in the counter of maximum Xenon ship producing stations depending on the alliance between the factions.
For example: if the Argon and Antigone are friends - the Xenons get +1 port and +1 shipyard; if the Paranid and the Holy Order are reconciled - the Xenons get another +1 shipyard and +1 port...
Similarly we could increase the coefficient for the maximum numberof ships available for construction also.
This way, we can get a little closer to parity: avoid the imbalance when 6-7 ports of one faction unsuccessfully try to resist 10+ ports of factions alliance.

This is not to mention that the Xenons simply can never win due to the presence of invulnerable objects in some factions (the problem is discussed in a separate thread)