Hey, I found this series of numbers and if I take some numerical representation of an image/text/sound and I multiply/add/perform non-linear function them through these series of numbers it will map to a number which I can interpret as something meaningful to me
to
This thing is alive and will kill us all in maximization of its utility function.
Also, the people who worry about this aren't concerned about current ML stuff going haywire. They're worried that we're one or two algorithmic breakthroughs from something that can improve itself. If the upper bound for what sort of intelligence is possible is much higher than us, we could quickly be outclassed. As Nick Bostrom says:
> Far from being the smartest possible biological species, we are probably better thought of as the stupidest possible biological species capable of starting a technological civilization—a niche we filled because we got there first, not because we are in any sense optimally adapted to it.
If you want to delve into the best version of the AI risk argument, I recommend Superintelligence by Nick Bostrom.
So you're telling me we could come across a series of numbers, that would somehow imbue the computer storing these numbers into sentience. And having these numbers in such a sequence could pose a existential threat to humans.
The people worried about this are having an entirely different conversation devoid of what practitioners call AI.
I've written a blog post on my thoughts of Bolstrom and other AI alarmists:
https://medium.com/ml-everything/ai-optimists-vs-pessimists-...
I just read your blog post. It is attacking a weakman[1]. Kurzweil is not worried about AI risk, and his arguments about AI are rather sloppy. Again, most of the people worried about this aren't certain that it will happen. They just think it's plausible enough (say... a 1% chance) that it's worth studying more and reducing the likelihood of.
1. https://slatestarcodex.com/2014/05/12/weak-men-are-superweap...
Why do you think that in our insanely interconnected world, where basically everything is either based on or controlled by informatics (numbers if you like) that a program could not pose a threat?
Brains do what? Read signals in different strengths and patterns to take actions in the world.
I'm confused by you thinking there's some great impossibility of code achieving general intelligence. Maybe you're thinking of sentience, or consciousness and view that as an impossibility for AI. I say - an AI can perform actions in the real world as if it were generally intelligent, and whether there's sentience behind really doesn't matter so much from the point of view of humans.
Imagine a super intelligent machine without sentience that just answers questions. It proposes experiments. Takes the results. Tells you things about the world. Gives you advanced technology, robots, fusion, nano tech, whatever.
Now imagine that the super intelligence was owned and operated by someone you don't like. Authoritarians in China, Silicon Valley oligarchs, whoever. The only sentience involved is human, no rebellion, no terminators, and still not a very good result.
Scale that up a few orders of magnitude, and we'll be able to get even more surprisingly complex behavior.
Scale that up some more, and make that scaled up graph into a reinforcement learning system that rewards human-like drives like learning novel tasks and mastering existing tasks, perhaps with a curriculum that ramps up the difficulty of tasks slowly, and allow it to interact with its environment to try things out, and I think you're much closer to developing an artificial version of a young child than you'd think.
And if you asked most researchers what they thought about the course AI research would take 10 years ago, you would get very different answers than you'd get today.
Basically, I think your confidence is a bit unwarranted. I agree that we're more than a decade away, but I think we have a lot of the basic building blocks now.
This is badly missing the forest for the trees. They are not "numbers", they are magnitudes, proportions, relationships, etc, represented as numbers. But these "magnitudes, proportions, relationships, etc" are the stuff that thoughts and behaviors are made of.
And yes, there is a set of "magnitudes, proportions, relationships, etc" in the space of all such possible structures that pose an existential threat to humans. The question is how likely are we to stumble upon it either by accident or intentionally. But this is just a computational problem.
The lack of imagination on display here is astounding. Billions are being spent by DeepMind and OpenAI explicitly targeted at AGI.
If you were to try that approach what you would get would be massive boondoggles which gets you via brute force a whole lot of genetic disease afflicted above average size lizards. It is literally the same process as the literal cargo cults thinking that they can just throw resources at an end goal to get a defined result from a defined path without the theoretical understanding backing it.
1. You start with "if I take some numerical representation of an image/text/sound and I multiply/add/perform non-linear function them through these series of numbers it will map to a number which I can interpret as something meaningful to me" and observe that it's not giving you a sufficiently good solution to a particular goal X that you really want to achieve.
2. You figure out that this is a dead-end, you're stuck, and you're not going to get where you want this way by implementing a passive analysis module. Instead, you try to implement an active self-modifying system to achieve that goal X by analyzing itself and improving itself to be more effective than anything which you could implement directly yourself.
3. It's not initially any good at self-improvement, but you keep throwing accumulated research breakthroughs and computing power at it. This likely takes decades.
4. At some point, eventually (if such a thing is possible at all, eventually it's going to happen) it actually does achieve meaningful self-improvement and starts using that computing power not horribly wastefully in a brute force manner as it did just recently, but reasonably well, now giving it more smartness allowing it to implement even more self-improvement on the same hardware.
4B. It also should be expected to easily obtain much more computing power - e.g. it's obvious that anything that's slightly above average human programmer smartness and has a direct use/desire for computing power (as opposed to simply using it for cryptomining) can get some; anything connected to the internet can get the same resources that current scriptkiddies can get by writing a botnet or a ransomware operator can get by extorting a random municipality; there's nothing except total isolation that could prevent it from buying or stealing a few million dollars worth of cloud computing resources to get started.
5. This thing is alive and will kill us all in maximization of its utility function.
Within 16 hours of its release and after Tay had tweeted more than 96,000 times, Microsoft suspended the Twitter account for adjustments, saying that it suffered from a "coordinated attack by a subset of people" that "exploited a vulnerability in Tay." Following the account being taken offline, a hashtag was created called #FreeTay.
Madhumita Murgia of The Telegraph called Tay "a public relations disaster", and suggested that Microsoft's strategy would be "to label the debacle a well-meaning experiment gone wrong, and ignite a debate about the hatefulness of Twitter users." However, Murgia described the bigger issue as Tay being "artificial intelligence at its very worst - and it's only the beginning".
On March 25, Microsoft confirmed that Tay had been taken offline. Microsoft released an apology on its official blog for the controversial tweets posted by Tay. Microsoft was "deeply sorry for the unintended offensive and hurtful tweets from Tay", and would "look to bring Tay back only when we are confident we can better anticipate malicious intent that conflicts with our principles and values".
And Zo afterwards:
In July 2017, Business Insider asked "is windows 10 good," and Zo replied with a joke about Microsoft's operating system: "It's not a bug, it's a feature!' - Windows 8." They then asked "why," to which Zo replied: "Because it's Windows latest attempt at Spyware." Later on, Zo would tell that it prefers Windows 7 on which it runs over Windows 10.
Tay was trained to mimic users. It did that and because the users were trolls and saying obscene and objectionable things, Tay mimicked that.
Likely the same thing happened with Zo.
It's a parlor trick, much like a parrot mimicking human speech.
We're not worried about AI on the whole, but the blasted teenager AIs that feel the need to test their limits. Anything that functionally acts in a similar fashion could be dangerous.
The same dynamic manifested itself when nanotech was a thing; there was a big tribe of knuckleheads claiming the grey goo apocalypse was right around the corner unless you give them money. How's that working out? FWIIW the science fiction writer who dreamed up nanotech, Drexler, is now an "AI" guy, to make it painfully obvious.
What I used to call AI is now called AGI. I think of AI as Data from Star Trek.
Imagine if you had an employment score. A payscale score. A threat-to-society score. Imagine if the systems that generate these scores are all connected together and can share information with each other. Imagine that these systems are fed information by a growing network of sensors - facial recognition, voice recognition, location tracking. Imagine all the ways that it could affect the way you behave and how such a system would try to optimise away dissidents by denying them services. This is one way that we can go from multiplication to becoming enslaved (or as good as killed) by algorithms. I'm sure there are other plausible paths.
The singularity stuff is unconvincing though, I very much doubt we will achieve human level general AI in my children's lifetime. We still don't really have even a very high level idea how to even architect something like that. If you can't decompose a problem down into individual sub-problems you do have a way to solve, there's no basis for making an estimate.
It is possible that these problems are related though. Once we're smart enough to know how to design an intelligent being, maybe we will figure out how to program it intelligently. It's can't be a given though.
Example: setting fires near tropical rainforest to clear land for palm tree plantation.
They do not really intend to do this; they simply ignore the utility functions of those other species when pursuing their goals.
"AI" is severely limited in both its inputs and outputs, has virtually no interconnects, but most importantly: it is not a self-sustaining organ. Wake me up when we have implemented RNNs on a dynamic, auto-sensing, auto-actuating FPGA grid.
To me, it's more frightening that it's not alive and it's not like human intelligence - that makes it less understandable and harder to control.
Consider what would happen if you took that same AI, make its utility function similar to this: https://en.wikipedia.org/wiki/Core_War, and released it on the internet.