3 ms·

I'm skeptical because we are building black boxes. How do you fix something you can't reason about? These billion parameter boxes are outside the reach of your

by mhitza 4y ago

I'm skeptical because we are building black boxes. How do you fix something you can't reason about? These billion parameter boxes are outside the reach of your everyday developers. In terms of cost of propping up the infrastructure makes them tenable only for megacorps. Most of us aren't moving goal posts, but are very much skeptic at the things we are being oversold on. I personally think we are still far away from AGI, and neural networks of any variety are converging on a local optima in the AI design space. I would enjoy "talking" with an AI that doesn't have the contextual memory of a proverbial gold fish. The real scary thing is that these objectively unprovable systems are plopped into existing systems and more and more in charge of automatic decision making. A corporation's wet dream, if they can absolve themselves of any responsibility "the algorithm can't lie!"

ngamboa 4y ago
Jyaif 4y ago
> I'm skeptical because we are building black boxes. Just want to point out that you are also a blackbox. And if you are going to say that you are not a blackbox because you can explain your reasoning, just know that some AIs already do that too.
- digitcatphd 4y ago
  To be fair, his point is you can't fix a black box and the human mind is still more a discipline of philosophy than modern science.
  - bradleykingz 4y ago
    Maybe we'll end up creating an artificial mind.
    - ben_w 4y ago
      I suspect we will. I hope we don’t give it e.g. a dark triad personality disorder when we do, though I fear we may — I suspect there more ways to make a broken mind than a healthy one.
rictic 4y ago
You're talking about a different sort of skepticism, about whether the effects of an AGI would be good or bad if one was produced with these methods. The skepticism that the parent comment was discussing was skepticism about whether we're on a path to AGI, for good or for ill.
Hellicio 4y ago
They are blackboxes for the normal user the same way as a smartphone is a blackbox. Non of my close family understands the technical detail from bits to an image. There are also plenty of expert systems were plenty of developers see them as blackboxes. Even normal databases and query optimizations are often enough blackboxes. As long as those systems perform better as existing systems, thats fine by me. Take auto pilot: As long as we can show/proofe good enough that it drives better than an untrained 18 year old or 80 year old (to take extremes, i'm actually quite an avg driver myself), all is good. And one very very big factor in my point of view: We never ever had the software equivilent of learning. When you look at Nvidia Omniverse, we are able to simulate those real life things so well, so often in such different scenarios, that we are already out of the loop. I can't drive 10 Million KM in my lifetime (i think). The cars from Google and Tesla already did that. Yesterday at the google io, they showed the big 50x Billion parameter network and for google this is the perfect excuse to gather and put all of this data they always had into something they now can monetarize. No one can ask google for money now like the press did (Same with Dall-E 2) I think its much more critical that we enforce/force corporations to make/keep those models free for everyone to use. unfortunate i have no clue how much hardware you need to run those huge models.
- esjeon 4y ago
  > They are blackboxes for the normal user the same way as a smartphone is a blackbox. You can't take that approach. The current NN techniques are blackbox-by-nature, and are blockbox to everyone including the devs. Proprietary software is only blackbox to consumers, and large complex software still have insides that can be observed when things go wrong. For NNs, nothing can describe how exactly they work, and each network has to be reverse engineered independently, which is, AFAIK, a separate research field. > I can't drive 10 Million KM in my lifetime (i think). The cars from Google and Tesla already did. The length (nor the amount of data) alone doesn't decide the quality of AI. Actually, ROI diminishes rather quickly in the early stage of development. The rest is about picking up corner cases. They can drive 1 parsec, and still would not perform better than the ones we have now. Also, again, because NN is a complete blackbox, even the devs can't be sure if those corner cases are properly reflected to a newly trained network, nor if the new training data didn't impact the performance in other corners. We just don't know for sure. We just take chances. That's the limitation.
uoaei 4y ago
> I'm skeptical because we are building black boxes. An article came up a couple days ago that points to some interpretable features of the so-called black boxes you refer to. It's not that they are black boxes, it's that our torches are not yet bright enough. https://vaclavkosar.com/ml/googles-pathways-language-model-and-chain-of-thought https://vaclavkosar.com/ml/googles-pathways-language-model-a... > Most of us aren't moving goal posts, but are very much skeptic at the things we are being oversold on. I think a shift in perspective is warranted here. It's becoming increasingly clear that we may have vastly overestimated our own intelligence. Human exceptionalism is crumbling before us as we see how limited the tools are that pull off such incredible stunts. Judging based on other neuroscience and psychology research coming out, it really does seem like we are no more than statistical inference machines with specialized hardware that allow us to pack a lot of processing power into a small, energy-efficient system. We need to figure out next better learning algorithms, which probably depend quite heavily on the particular physical architecture.

by mhitza 4y ago

ngamboa 4y ago

Jyaif 4y ago

> I'm skeptical because we are building black boxes. Just want to point out that you are also a blackbox. And if you are going to say that you are not a blackbox because you can explain your reasoning, just know that some AIs already do that too.

digitcatphd 4y ago
To be fair, his point is you can't fix a black box and the human mind is still more a discipline of philosophy than modern science.
- bradleykingz 4y ago
  Maybe we'll end up creating an artificial mind.
  - ben_w 4y ago
    I suspect we will. I hope we don’t give it e.g. a dark triad personality disorder when we do, though I fear we may — I suspect there more ways to make a broken mind than a healthy one.

rictic 4y ago

You're talking about a different sort of skepticism, about whether the effects of an AGI would be good or bad if one was produced with these methods. The skepticism that the parent comment was discussing was skepticism about whether we're on a path to AGI, for good or for ill.

Hellicio 4y ago

They are blackboxes for the normal user the same way as a smartphone is a blackbox. Non of my close family understands the technical detail from bits to an image. There are also plenty of expert systems were plenty of developers see them as blackboxes. Even normal databases and query optimizations are often enough blackboxes. As long as those systems perform better as existing systems, thats fine by me. Take auto pilot: As long as we can show/proofe good enough that it drives better than an untrained 18 year old or 80 year old (to take extremes, i'm actually quite an avg driver myself), all is good. And one very very big factor in my point of view: We never ever had the software equivilent of learning. When you look at Nvidia Omniverse, we are able to simulate those real life things so well, so often in such different scenarios, that we are already out of the loop. I can't drive 10 Million KM in my lifetime (i think). The cars from Google and Tesla already did that. Yesterday at the google io, they showed the big 50x Billion parameter network and for google this is the perfect excuse to gather and put all of this data they always had into something they now can monetarize. No one can ask google for money now like the press did (Same with Dall-E 2) I think its much more critical that we enforce/force corporations to make/keep those models free for everyone to use. unfortunate i have no clue how much hardware you need to run those huge models.

esjeon 4y ago
> They are blackboxes for the normal user the same way as a smartphone is a blackbox. You can't take that approach. The current NN techniques are blackbox-by-nature, and are blockbox to everyone including the devs. Proprietary software is only blackbox to consumers, and large complex software still have insides that can be observed when things go wrong. For NNs, nothing can describe how exactly they work, and each network has to be reverse engineered independently, which is, AFAIK, a separate research field. > I can't drive 10 Million KM in my lifetime (i think). The cars from Google and Tesla already did. The length (nor the amount of data) alone doesn't decide the quality of AI. Actually, ROI diminishes rather quickly in the early stage of development. The rest is about picking up corner cases. They can drive 1 parsec, and still would not perform better than the ones we have now. Also, again, because NN is a complete blackbox, even the devs can't be sure if those corner cases are properly reflected to a newly trained network, nor if the new training data didn't impact the performance in other corners. We just don't know for sure. We just take chances. That's the limitation.

uoaei 4y ago

> I'm skeptical because we are building black boxes. An article came up a couple days ago that points to some interpretable features of the so-called black boxes you refer to. It's not that they are black boxes, it's that our torches are not yet bright enough. https://vaclavkosar.com/ml/googles-pathways-language-model-and-chain-of-thought https://vaclavkosar.com/ml/googles-pathways-language-model-a... > Most of us aren't moving goal posts, but are very much skeptic at the things we are being oversold on. I think a shift in perspective is warranted here. It's becoming increasingly clear that we may have vastly overestimated our own intelligence. Human exceptionalism is crumbling before us as we see how limited the tools are that pull off such incredible stunts. Judging based on other neuroscience and psychology research coming out, it really does seem like we are no more than statistical inference machines with specialized hardware that allow us to pack a lot of processing power into a small, energy-efficient system. We need to figure out next better learning algorithms, which probably depend quite heavily on the particular physical architecture.