Is everything a religion?

[Kidlicious]

Tht means the following:
The initial setup of your neural network (i.e. the intrconnectivity oof neurons) is determind by your genes, also certain characteristics (for example the density of D2 recepors, as I said)
But due to experiences (= learning) the strength of your neural connection get altered ... for example by increasing the amount of postsynaptic receptors or by a neurons and an axon forming more synapses.

Everyones decisions (including yours) are made due to the chemical reactions in his PNS and CNS (or, simpler put, due to the internal state of his state machine (current neural activity and current neurohormonal levels) and the external input from his sensors (i.e. eyes, ears, nose and so on). But due to the invivisual setup of his brain everyones reactions to the same stimuli may be different.
(example: Some persons reation to mild levels of pain may be sexual arousal ... most persons level to pain may be an avoidance reaction)

OK. I get what you are saying. But, like I was saying, I smoked for 20 years because of addiction. I didn't get off sexually on nicotine fits or anything. When I got a craving I smoked as soon as possible, just like other addicts. I actually don't know anyone who likes nicotine fits. I don't know if that's relevant.

I also edited my post (DanS you). What do you say about an alcoholic who has a predisposition to addiction (other family members are alcoholic) who quits. These people exist. You can find them at AA.

[Jon Miller]

I am suspicious that when we develop the needed technology to be able to handle open tasks (like in the real world, ones that don't lend themselves to models/loss functions) we will discover that it is hard to go from dog level intelligence to human, and that it might take a long time to go from child to adult.

I think the 'singularity' is much more likely to be based on biological minds and a mind/machine interface that allows biological minds to fully exploit modeling and loss functions which we can exploit with difficulty now.

jm

[Kidlicious]

I am suspicious that when we develop the needed technology to be able to handle open tasks (like in the real world, ones that don't lend themselves to models/loss functions) we will discover that it is hard to go from dog level intelligence to human, and that it might take a long time to go from child to adult.

I think the 'singularity' is much more likely to be based on biological minds and a mind/machine interface that allows biological minds to fully exploit modeling and loss functions which we can exploit with difficulty now.

jm

When you say "human level intelligence" what do you mean exactly? Do you mean that they will have their own will? Because dogs and children have that.

[Aeson]

I am suspicious that when we develop the needed technology to be able to handle open tasks (like in the real world, ones that don't lend themselves to models/loss functions) we will discover that it is hard to go from dog level intelligence to human, and that it might take a long time to go from child to adult.

I think the 'singularity' is much more likely to be based on biological minds and a mind/machine interface that allows biological minds to fully exploit modeling and loss functions which we can exploit with difficulty now.

jm

At the point we get dog level AI, we'll have already built what is necessary for it to quickly exceed us (at least our natural capabilities). This is because to get to dog level we'll need real learning algorithms with applicable self-modification (so simulated evolution). At that point it's just a matter generations and how long they take. Likely they happen very fast.

[Jon Miller]

At the point we get dog level AI, we'll have already built what is necessary for it to quickly exceed us (at least our natural capabilities). This is because to get to dog level we'll need real learning algorithms with applicable self-modification (so simulated evolution). At that point it's just a matter generations and how long they take. Likely they happen very fast.

That is likely to happen really fast for a model and a loss function.

I am suspicious that for what is needed to handle an open environment, it won't be really fast... but will be slow even for much greater resources (and we are running to the end of moores law anyway).

Of course, you do point to 'natural capabilities'. To be able to naturally, instead of via keyboard interface, take advantage of our model/loss function based tools that we use on computers will be a huge step forward. But this is going to be done much earlier by 'born humans' then it is by AI (which needs some development of something that we have no idea of how to do/etc and so could take a really really long time).

JM
(For example, we have known that quantum mechanics + general relativity need a new theory since the 20s.... we are just coming to the conclusion now that we have not really made any real progress on that front, despite the increasingly rapid development of a lot of really great theories)

[Jon Miller]

We already have monkey brains controlling robot arms. Mind machine interfacing is coming much quicker then AI.

JM

[Aeson]

I agree augmentation (memory and computation) is likely to happen first. At which point humans become much more likely to be able to develop true AI. The interface would require and/or expose a lot of applicable information.

Perhaps augmentation will eliminate the need for AI at all, with expert systems added onto human intelligence being able to handle everything we think of. Though from a competition/military standpoint it will always make sense to be the first to get there.

Hopefully by then humanity has moved beyond competing/military.

[Lorizael]

I think a lot of talk about what constitutes "real" AI conflates sentience, intelligence, and free will. While those concepts are clearly linked in terms of what we observe--humans--we have enough science fiction out there to imagine something different--that is, to see them as distinct concepts that don't necessarily imply each other. It might turn out that the concepts are inextricably, fundamentally linked, but I don't think we're at all knowledgeable enough yet to say that they are.

[Broken_Erika]

Hopefully by then humanity has moved beyond competing/military.

Don't count on it, there will probably be a world war XXX.

[Kidlicious]

Don't count on it, there will probably be a world war XXX.

Not if we just get used to terrorist attacks.

[Elok]

Don't count on it, there will probably be a world war XXX.

Well, all we have to do is remove all resource scarcity and/or eliminate population growth entirely. No biggie.

[Proteus_MST]

Well, all we have to do is remove all resource scarcity and/or eliminate population growth entirely. No biggie.

Then we would still have wars for reasons that can be attributed to different races or religions (and intra societal violence because some groups cannot accept other groups sexual orientation)

[Broken_Erika]

Eliminate religion and you'll have atheist groups fighting each other over who has the most logical name "Atheists United" or "United Atheists"

[Kidlicious]

Eliminate religion and you'll have atheist groups fighting each other over who has the most logical name "Atheists United" or "United Atheists"

... and calling each other religionists.

Great job getting the thread back on track.

[Elok]

Then we would still have wars for reasons that can be attributed to different races or religions (and intra societal violence because some groups cannot accept other groups sexual orientation)

A truly post-scarcity world (while grossly improbable) would eliminate most such conflicts. American race troubles, for example, are heavily rooted in economic concerns. We drove the Indians off their land because they used it inefficiently and we had a lot of people looking to settle it. Then we imported masses of African slaves because that happened to be the cheapest source of agricultural labor. Now we hate Mexicans because they're used to circumvent labor laws and depress wages. And our screwy relationship with Arabs is mostly about our heavy-handed methods to secure access to their petroleum. All four groups are now heavily disadvantaged, which helps them nurse an entirely rational grudge.

Religious/cultural conflicts are more complicated, because those can hide the real issues beneath rhetoric and apparently abstruse arguments. Even then, I would argue that most theological spats were ultimately about political power. A lot of early Christian controversies, for example, were about battles for influence between the Patriarch of Constantinople and the older bishoprics of places like Antioch and Alexandria. Later in the Middle Ages, things like the Investiture Controversy concerned the extent of Papal power. Nobody would have much cared about the Albigensians if they didn't represent an effective secession movement from the power base of the RCC. Etc.