10 ms·

How should we evaluate progress in AI?

119 points by amasad 8y ago

tim333 8y ago
This seems to an article written by a philosophical type who doesn't really understand contemporary AI from a technical point of view. I'm not sure it's terribly useful.
- rrherr 8y ago
  About the author: “I did a PhD in artificial intelligence at MIT. My undergraduate degree was in math. I’ve also studied cognitive science, biochemistry, Old English and Ancient Greek literature. None of that qualifies me to write Meaningness, but it may explain a certain STEM-ish orientation, decorated with occasional literary jokes. ... I have founded, managed, grown, and sold a successful biotech informatics company. That may explain a certain practical orientation, and lack of interest in philosophical theories that depend on the world being very unlike the way it appears.” https://meaningness.com/about-my-sites https://meaningness.com/about-my-sites
  - tim333 8y ago
    Ah maybe not then.
    - nyolfen 8y ago
      the site this blog is hosted on is really great for explaining a lot of aspects of postmodernism to engineering-oriented or analytical thinkers. i got totally sucked in by it over my christmas vacation two years ago and read the whole thing.
    - montenegrohugo 8y ago
      I loved this Sorry tim! :P
  - visarga 8y ago
    I expected better insights from a person who has a PhD in AI and spans multiple other fields. I think there is a gap between philosophy and AI, and philosophy has a lot of catching up to do. I'm especially looking at the "hard problem" and "qualia". In the meantime AI people implemented a lot of the activities of the mind and philosophers are still stuck on these toy concepts that don't go so far. Instead, they should focus on the implications of agent-environment learning - reinforcement learning and evolutionary strategies - as the true sources of human intelligence.
mlthoughts2018 8y ago
> “It’s not scientific progress unless you understand where the improvement is coming from.” I don’t agree with this. If you can chronicle improvement, that is progress. Giving a satisfying linguistic description of that improvement, when possible, might be more progress, but merely documenting it is extremely important scientific progress in its own right. Overall this essay was extremely hard to read and should cut down about 75% of the content. The whole wolpertinger thing is nothing but a distraction. Just say AI is a mixture of disciplines and serves a mixture of outcomes. It only takes away from the essay to act like you’re being literary or nuanced with the wolpertinger thing when all it does is subtract from the arguments. And to boot, after so many words, the final advice is extremely hollow... literally just saying, > “And so we should try to do better along lots of axes.” How should we improve? I guess by “doing better” on “multiple axes.” The section on “antidotes” is hardly better, saying: > “I will suggest two antidotes. The first is the design practice of maintaining continuous contact with the concrete, nebulous real-world problem. Retreating into abstract problem-solving is tidier but usually doesn’t work well.“ Except this is already what basically everyone tries to do. Research labs try to maintain direct contact with state of the art benchmark tasks on a wide variety of data sets. And often they work extremely hard to produce results robust across several tasks and several data sets. And in various other fractured or specific cases, the researchers are very clear up-front they are solving one particular, ultraspecific problem in the scope of the paper. (Unfortunately the second antidote is more “wolpertinger”... ugh.)
- amasad 8y ago
  > And to boot, after so many words, the final advice is extremely hollow... literally just saying, > > “And so we should try to do better along lots of axes.” > How should we improve? I guess by “doing better” on “multiple axes.” That's not what the final advice is, the author is suggesting the use of "meta-rationality": > "AI is a wolpertinger: not a coherent, unified technical discipline, but a peculiar hybrid of fields with diverse ways of seeing, diverse criteria for progress, and diverse rational and non-rational methods. Characteristically, meta-rationality evaluates, selects, combines, modifies, discovers, creates, and monitors multiple frameworks." Although not expanded on in this essay, it seems like the whole blog is dedicated to the topic.
andyidsinga 8y ago
As a person that works in software development and trying to make sense of data - I found the article pretty interesting after the paragraph below. The discussion around science, engineering and adjacencies are good to keep in mind. > AI researchers often say they are doing engineering. This can sound defensive, when you point out that they aren’t doing science: “Yeah, well, I’m just doing engineering, making this widget work better.” It can also sound derisive, when you suggest that philosophical considerations are relevant: “I’m doing real work, so that airy-fairy stuff is irrelevant. As an engineer, I think metaphysics is b.s.”
- aisofteng 8y ago
  Any engineer that thinks metaphysics is “b.s.” is uneducated in the schools of thought that made engineering of any sort possible in the first place.
  - andyidsinga 8y ago
    yep -- sadly, I've heard that sentiment over and over. edit: I should clarify - I've heard that sentiment a lot, but now that I think of it, the "it's BS" part usually comes from a place of frustration due to time constraint. ..which leads to this article about why software is late: https://www.computer.org/cms/Computer.org/ComputingNow/homepage/2011/1111/W_SW_AllLateProjectsAretheSame.pdf https://www.computer.org/cms/Computer.org/ComputingNow/homep...
  - j88439h84 8y ago
    Can you explain what you mean by this?
- joe_the_user 8y ago
  Well, one interesting thing about the article is the way it describes GOFAI as based on false ideas, which is harsh but fair. But in a lot of ways, GOFAI was "applied philosophy", at least something of a direct map of one form of rationalist philosophy to computation. GOFAI was abandoned but the various perspectives remain and current AI goes forward without necessarily having any larger perspective at all, just "make it work". One might guess we're waiting for a different coherent philosophy to organize the process.
joejerryronnie 8y ago
I am not trained in any type of field that is remotely related to AI research or engineering. Outside of some basic ML projects at work, I am not well versed in the practical application of AI technologies. But I do wonder a few things about AI: - Have we made real, technological progress over the last 50 years or are we just leveraging far greater computing power and the ability to collect much larger data sets to run statistical analysis on? - Will general purpose AI consist of essentially layers and layers of AI's that can handle progressively more abstract inputs, models, and patterns? For instance, the lowest level AI is what we see today - a very powerful tool but bound to a specific use case. One layer up may be able to combine inputs from a dozen 1st tier AI's to generalize a tiny bit more on the individual use cases and can deal with a tiny bit more ambiguity. One level up will evaluate inputs from a dozen level 2 AI's and so on. With the final top layer (perhaps millions of levels up) resembling general purpose processing similar to a human brain. What if this model ended up producing true general purpose AI, but the amount of input synthesizing and modeling required so much processing power that the speed at which general purpose AI could operate is no faster than a human brain? - Can we achieve general purpose AI through purely algorithmic means, or will we need to implement a hybrid biological model to achieve real breakthroughs? If we could accomplish this, would we understand the detailed mechanisms of the biological component of the hybrid or would it forever remain a black box that we just tap into? Anyway, not sure this adds a whole lot to the specific discussion on how best to measure AI progress, but they're questions I've been pondering lately.
- pnloyd 8y ago
  Ya that's kind of an interesting question. With Moore's law starting to approach it's physical limitations it would seem AGI wouldn't be feasible with current algorithms. Not to mentions as your describing.. those "millions" of layers of narrow AI's sounds like impossible amount of work to do.. I don't think very many machine learning experts really believe that those techniques will lead to AGI.
  - taeric 8y ago
    I'm curious just how true it is that Moore's law is starting to approach physical limitations. I just recently listened to some of Feynman's speeches collected in [1]. One of them was about how to place all of the works of an encyclopedia onto the tip of a pin. Did it cover anything we couldn't do today? No. But that was the point. Just using a somewhat naive view of the physical matter that you would be putting something on, it was possible to go quite dense. Imagine if we started going even denser. Do I suspect we are approaching limits? Certainly. Question is more of just how much further we can go. And will we need a dramatic shift of any sort before we could realize some extra distance? [1] https://www.audible.com/pd/Science-Technology/The-Pleasure-of-Finding-Things-Out-Audiobook/B00BSU83HI https://www.audible.com/pd/Science-Technology/The-Pleasure-o...
taeric 8y ago
I confess I was tempted to take a pass on this one. In large, because I've fatigued a lot on reading about AI and Machine Learning. Didn't help that this is a large article. That said, I encourage everyone to give this more than a single pass. There is irony that we are, to this day, still quoting and agreeing with Feynman's Cargo Cult Science piece. It is almost disheartening to see that we still have a hard time listening to advice from the early 1970s. By and large, though, this piece does a great job really laying out what makes it so hard to level most criticisms at AI related studies. The cross disciplinary look is one I wouldn't have thought to do, but really does explain a lot. I'm torn, because I'm sympathetic to most of the defenses. It is hard to really believe we are making meaningful progress, though. Even if I am enjoying many of the small practical improvements we have managed to get out of things.
- Erlich_Bachman 8y ago
  What would "meaningful progress" entail for you? How would your life change, how would the life on earth in general change? Great achievements in science (aided by AI)?
  - taeric 8y ago
    That is a different question, though. The question if if there is meaningful progress in ai. The amount of progress explained solely by percentage improvements against a benchmark is pretty high. I don't think this is particularly damning. Nor do I think it should be halted. However, I agree it is hard to call progress.
    - Erlich_Bachman 8y ago
      What specifically would be easy for you to call progress?
      - taeric 8y ago
        An equation that could predict a better ML model. Honestly, I think claiming I want an equation is a touch too much. However, a falsifiable prediction would be nice. Imagine if the only way we got accurate ballistics was by requiring faster more powerful guns all of the time. "We can hit the target, but only if we upgrade our guns to railguns and limit ourselves to large targets."
jmickey 8y ago
Relevant - AI Progress Measurement page by the EFF: https://www.eff.org/ai/metrics https://www.eff.org/ai/metrics
toolslive 8y ago
For typical problems (playing chess, tissue segmentation, translation, ...) you have the quality of the solution (sometimes difficult to measure) versus the cost of achieving it (energy/enthropy and maybe time). The cost is important.
stared 8y ago
Please, no. While he provides some food for thought, asks a few important questions (and provokes even more), this text is full of pseudointellectualism. As in: fancy words but utter lack of understanding of the core subjects one writes about. Most importantly: what is science (no, it is not a trivial question; I recommend going Ludwik Fleck's "Genesis and development of a scientific fact" route, http://www.evolocus.com/Textbooks/Fleck1979.pdf http://www.evolocus.com/Textbooks/Fleck1979.pdf) and what is AI (it is a vast field; some of it IS math/CS (as in: proving things), but it is a small part). For example, many of things he says would fit other practical disciplines, e.g. medicine. Yes, experiments are on a group of people. Yes, criterion that "drug X works better than drug Y, but we don't know why" is sufficient. > I don’t know data science folks well, but my impression is that they find the inexplicability and unreliability of AI methods frustrating. Well, it is not the main issue (speaking as a data scientist working with AI). At least he acknowledges his lack of expertise. > These failures of scientific practice seem as common in AI research now as they were in social psychology a decade ago. From psychology’s experience, we should expect that many supposed AI results are scientifically false. Also - no, it is not at level of psychology when it comes to the replication crisis. A lot of code (though, unfortunately, not all) is shared online, by the authors or other contributors, and people do replicate it (or there is an absence of replication, which also conveys a message).
- Radim 8y ago
  Speaking as both a ML researcher and an applied ML business owner (one who hired stared at one point — hi Piotr :), I respectfully disagree. The replication crisis in "AI" may not be as bad as psychology (I wouldn't know), but it's not great. Sadly, my brain has somehow learned to equate "SOTA" with "hot-stitched crap, stay away". Too many painful lessons. On the subject of publishing code: this is useful to the degree that it removes bad faith as the possible reason for the lack of replicability. But otherwise helps little in practical terms. You just have the privilege to sift through the bugs and bad design in more close-up. "I am afraid you are right. I used to reach ~72% via the given random seed on an old version of pytorch, but now with the new version of pytorch, I wasn't able to reproduce the result. My personal opinion is that the model is neither deep or sophisticated, and usually for such kind of model, tuning hyper parameters will change the results a lot (although I don't think it's worthy to invest time tweaking an unstable model structure)." = quote [1] for one of the "new SOTA" papers from NLP (WikiQA question answering), where the replication scores came out 62% instead of claimed 72%. I generally call this the "AI Mummy Effect" — looks great but crumbles to dust on touch. [1] https://github.com/pcgreat/SeqMatchSeq/issues/1 https://github.com/pcgreat/SeqMatchSeq/issues/1
sgt101 8y ago
If the subject is seen as a system rather than a thing you can make some progress. On the one hand we have Artificial Intelligence, where insights from cognition and biology are used to model and explain reasoning and behaviour. On the other we have AI where people use the outcome of Artificial Intelligence research with other pragmatically selected technical components to develop technology. I think that there is interdependence and exchange between the two, but they have different methodologies and processes. I also think that huge trouble is created when members of one community use the other communities clothes and achievements. For example Artificial Intelligence researchers talking about the practical impact of AI and near term application of their work and conversely AI researchers claiming that their work "works like a brain".
jfjfjfjfyy 8y ago
When AI progresses, it evaluates you, not the other way round. By the time this is realised, it is already too late for yet another 'generation' of AI researchers.
laichzeit0 8y ago
> “This year, we’re getting Z% correct, whereas last year we could only get (Z-ε)%” does sound like progress. But is it meaningful? This is one thing that bother's me a lot when I read published work. I have a feeling this is a result of everyone using the same benchmark datasets, so it inevitably becomes more of an _engineering_ exercise rather than scientific progress. In NLP the difference between a publishable result and one that is not is often to squeeze out a few extra (Z-ε)% by throwing in an attention mechanism and ensembles to your new super duper improved SOTA architecture. This is the problem of "replicability", which really requires more than just the same benchmark dataset used over and over again. The author seems to touch on this point later on. Then there's the issue of "reproducibility". Very few researchers seem to publish their code with instructions how to build and re-create their results. What an awful lot of time is wasted trying to reproduce results. Here's a good example: https://groups.google.com/forum/#!topic/word2vec-toolkit/Q49FIrNOQRo https://groups.google.com/forum/#!topic/word2vec-toolkit/Q49...
ggm 8y ago
Most (human) languages have a set of rules and a variant set of variances ("i before e except after c, and this catalog of things which we pretend don't exist") so ML and NLP are cases of things which are ameanable to rules based systems, and because english corpus exists widely can be tested openly, against each other and a common norm of comprehension (for english speakers) generalized AI does not lie here: systems which uncover the grammer rules and exceptions do not generalize to systems which uncover rules in Law, or Equity, or financial trading, or other things. Yes, you can train nets. But the commonality here, is you can train, not that emergent AI is found. (not an AI person, strongly anti-AI perspective from life experience in compsci)
- fny 8y ago
  Is there such a thing as AGI really though instead of collections of modules that are trained and may have no meaningful mode of interaction? What does NLP have to do with the mathematics of financial trading? AGI has always felt to me as a large scale interdisciplinary/intercomputational activity, much in the same way that most human intelligence derived from years of intergenerational and interpersonal intellectual development. AGI will never be one but many: many intelligences and systems interacting to produce something of utlitarian value.
  - ggm 8y ago
    much in the same way that most human intelligence derived from years of intergenerational and interpersonal intellectual development. wow! I had no idea the archeology on early hominids was that good... (yes, implicit /s)
- taeric 8y ago
  This treats human language as a somewhat static thing that can be fully specified. Which, interestingly, avoids a large portion of human languages in poetry and metaphor. A bit like judging story telling by only accounting true stories, thus completely ignoring all of fiction. Saying nothing of false stories purported as true. It is quite possible that understanding is nothing more than a statistical relating of otherwise unrelated things. Some things, we accept as being 100% related. But, quite a lot of things are just "highly related." Where "highly" is probably not specified.
YeGoblynQueenne 8y ago
>> GOFAI had several defects, but… the main thing is, nearly all of it was false. It would be nice to see some kind of substantial examples of how "(nearly all of) GOFAI was false", accompanying statements like the one above. The problem of course is - those are very hard to come by. That is so because logic-based AI was abandoned. And it was abandoned because funding was cut repeatedly, not because of its failure to prove this theory or achieve that aim, but because the ones holding the purse strings were administrators and military pencil-pushers, who had no way to know a successful, or failed, program if one came up and bit them in the boogies. And just to substantiate my comment- what, exactly, was "false" about logic programming, one of the major research subjects in GOFAI? It worked just fine back then, it works just fine right now. In very practical, down to earth terms, you can prove a proposition, or a predicate, true or false by automatic means, sure as you can answer "2 + 2 = ?". So, really- more substance, less assertiveness, would do a world of good to those for whom "AI" means everything that they read online after 2012 and who may end up missing a hell of a lot of the history of the field if they take that sort of "GOFAI failed" statements at face value.
- artilect 8y ago
  Aigo.AI has spent a few million attempting to build AGI with a GOFAI architecture
- colorint 8y ago
  As a general matter, you can't prove a predicate true or false by automatic means. Logic programming is just as "artful" as imperative or functional programming, because they all run into the same problem: it's impossible to tell the difference between long-running and infinite computation. The question of algorithms for general logic was explicitly addressed as the Entscheidungsproblem, meaning decision problem, which was independently proven undecidable by both Turing and Church: https://en.wikipedia.org/wiki/Entscheidungsproblem https://en.wikipedia.org/wiki/Entscheidungsproblem
  - YeGoblynQueenne 8y ago
    >> As a general matter, you can't prove a predicate true or false by automatic means. Not in the general case, sure, yet in practice I'm sure we've all written plenty of code that terminates just fine. [Edit: I'm talking about imperative as well as logic or functional programming code]. The question then is- what does it mean when a program terminates? In the case of principled approaches like logic or functional programming, you have a pretty good idea what that means (e.g. a logic program proves a theory true or false). When an imperative program terminates, it's a very hairy affair to say what, exactly, termination means. [Edit 2: Actually, if you think about it, there's nothing we can really achieve in the general case (including machine learning; see language learning in the limit). In practice, on the other hand, we're doing things, alright - by continuously relaxing principles and fudging limits as necessary (see PAC learning)].

119 points by amasad 8y ago

This seems to an article written by a philosophical type who doesn't really understand contemporary AI from a technical point of view. I'm not sure it's terribly useful.

rrherr 8y ago
About the author: “I did a PhD in artificial intelligence at MIT. My undergraduate degree was in math. I’ve also studied cognitive science, biochemistry, Old English and Ancient Greek literature. None of that qualifies me to write Meaningness, but it may explain a certain STEM-ish orientation, decorated with occasional literary jokes. ... I have founded, managed, grown, and sold a successful biotech informatics company. That may explain a certain practical orientation, and lack of interest in philosophical theories that depend on the world being very unlike the way it appears.” https://meaningness.com/about-my-sites https://meaningness.com/about-my-sites
- tim333 8y ago
  Ah maybe not then.
  - nyolfen 8y ago
    the site this blog is hosted on is really great for explaining a lot of aspects of postmodernism to engineering-oriented or analytical thinkers. i got totally sucked in by it over my christmas vacation two years ago and read the whole thing.
  - montenegrohugo 8y ago
    I loved this Sorry tim! :P
- visarga 8y ago
  I expected better insights from a person who has a PhD in AI and spans multiple other fields. I think there is a gap between philosophy and AI, and philosophy has a lot of catching up to do. I'm especially looking at the "hard problem" and "qualia". In the meantime AI people implemented a lot of the activities of the mind and philosophers are still stuck on these toy concepts that don't go so far. Instead, they should focus on the implications of agent-environment learning - reinforcement learning and evolutionary strategies - as the true sources of human intelligence.

mlthoughts2018 8y ago

> “It’s not scientific progress unless you understand where the improvement is coming from.” I don’t agree with this. If you can chronicle improvement, that is progress. Giving a satisfying linguistic description of that improvement, when possible, might be more progress, but merely documenting it is extremely important scientific progress in its own right. Overall this essay was extremely hard to read and should cut down about 75% of the content. The whole wolpertinger thing is nothing but a distraction. Just say AI is a mixture of disciplines and serves a mixture of outcomes. It only takes away from the essay to act like you’re being literary or nuanced with the wolpertinger thing when all it does is subtract from the arguments. And to boot, after so many words, the final advice is extremely hollow... literally just saying, > “And so we should try to do better along lots of axes.” How should we improve? I guess by “doing better” on “multiple axes.” The section on “antidotes” is hardly better, saying: > “I will suggest two antidotes. The first is the design practice of maintaining continuous contact with the concrete, nebulous real-world problem. Retreating into abstract problem-solving is tidier but usually doesn’t work well.“ Except this is already what basically everyone tries to do. Research labs try to maintain direct contact with state of the art benchmark tasks on a wide variety of data sets. And often they work extremely hard to produce results robust across several tasks and several data sets. And in various other fractured or specific cases, the researchers are very clear up-front they are solving one particular, ultraspecific problem in the scope of the paper. (Unfortunately the second antidote is more “wolpertinger”... ugh.)

amasad 8y ago
> And to boot, after so many words, the final advice is extremely hollow... literally just saying, > > “And so we should try to do better along lots of axes.” > How should we improve? I guess by “doing better” on “multiple axes.” That's not what the final advice is, the author is suggesting the use of "meta-rationality": > "AI is a wolpertinger: not a coherent, unified technical discipline, but a peculiar hybrid of fields with diverse ways of seeing, diverse criteria for progress, and diverse rational and non-rational methods. Characteristically, meta-rationality evaluates, selects, combines, modifies, discovers, creates, and monitors multiple frameworks." Although not expanded on in this essay, it seems like the whole blog is dedicated to the topic.

andyidsinga 8y ago

As a person that works in software development and trying to make sense of data - I found the article pretty interesting after the paragraph below. The discussion around science, engineering and adjacencies are good to keep in mind. > AI researchers often say they are doing engineering. This can sound defensive, when you point out that they aren’t doing science: “Yeah, well, I’m just doing engineering, making this widget work better.” It can also sound derisive, when you suggest that philosophical considerations are relevant: “I’m doing real work, so that airy-fairy stuff is irrelevant. As an engineer, I think metaphysics is b.s.”

aisofteng 8y ago
Any engineer that thinks metaphysics is “b.s.” is uneducated in the schools of thought that made engineering of any sort possible in the first place.
- andyidsinga 8y ago
  yep -- sadly, I've heard that sentiment over and over. edit: I should clarify - I've heard that sentiment a lot, but now that I think of it, the "it's BS" part usually comes from a place of frustration due to time constraint. ..which leads to this article about why software is late: https://www.computer.org/cms/Computer.org/ComputingNow/homepage/2011/1111/W_SW_AllLateProjectsAretheSame.pdf https://www.computer.org/cms/Computer.org/ComputingNow/homep...
- j88439h84 8y ago
  Can you explain what you mean by this?
joe_the_user 8y ago
Well, one interesting thing about the article is the way it describes GOFAI as based on false ideas, which is harsh but fair. But in a lot of ways, GOFAI was "applied philosophy", at least something of a direct map of one form of rationalist philosophy to computation. GOFAI was abandoned but the various perspectives remain and current AI goes forward without necessarily having any larger perspective at all, just "make it work". One might guess we're waiting for a different coherent philosophy to organize the process.

joejerryronnie 8y ago

I am not trained in any type of field that is remotely related to AI research or engineering. Outside of some basic ML projects at work, I am not well versed in the practical application of AI technologies. But I do wonder a few things about AI: - Have we made real, technological progress over the last 50 years or are we just leveraging far greater computing power and the ability to collect much larger data sets to run statistical analysis on? - Will general purpose AI consist of essentially layers and layers of AI's that can handle progressively more abstract inputs, models, and patterns? For instance, the lowest level AI is what we see today - a very powerful tool but bound to a specific use case. One layer up may be able to combine inputs from a dozen 1st tier AI's to generalize a tiny bit more on the individual use cases and can deal with a tiny bit more ambiguity. One level up will evaluate inputs from a dozen level 2 AI's and so on. With the final top layer (perhaps millions of levels up) resembling general purpose processing similar to a human brain. What if this model ended up producing true general purpose AI, but the amount of input synthesizing and modeling required so much processing power that the speed at which general purpose AI could operate is no faster than a human brain? - Can we achieve general purpose AI through purely algorithmic means, or will we need to implement a hybrid biological model to achieve real breakthroughs? If we could accomplish this, would we understand the detailed mechanisms of the biological component of the hybrid or would it forever remain a black box that we just tap into? Anyway, not sure this adds a whole lot to the specific discussion on how best to measure AI progress, but they're questions I've been pondering lately.

pnloyd 8y ago
Ya that's kind of an interesting question. With Moore's law starting to approach it's physical limitations it would seem AGI wouldn't be feasible with current algorithms. Not to mentions as your describing.. those "millions" of layers of narrow AI's sounds like impossible amount of work to do.. I don't think very many machine learning experts really believe that those techniques will lead to AGI.
- taeric 8y ago
  I'm curious just how true it is that Moore's law is starting to approach physical limitations. I just recently listened to some of Feynman's speeches collected in [1]. One of them was about how to place all of the works of an encyclopedia onto the tip of a pin. Did it cover anything we couldn't do today? No. But that was the point. Just using a somewhat naive view of the physical matter that you would be putting something on, it was possible to go quite dense. Imagine if we started going even denser. Do I suspect we are approaching limits? Certainly. Question is more of just how much further we can go. And will we need a dramatic shift of any sort before we could realize some extra distance? [1] https://www.audible.com/pd/Science-Technology/The-Pleasure-of-Finding-Things-Out-Audiobook/B00BSU83HI https://www.audible.com/pd/Science-Technology/The-Pleasure-o...

taeric 8y ago

I confess I was tempted to take a pass on this one. In large, because I've fatigued a lot on reading about AI and Machine Learning. Didn't help that this is a large article. That said, I encourage everyone to give this more than a single pass. There is irony that we are, to this day, still quoting and agreeing with Feynman's Cargo Cult Science piece. It is almost disheartening to see that we still have a hard time listening to advice from the early 1970s. By and large, though, this piece does a great job really laying out what makes it so hard to level most criticisms at AI related studies. The cross disciplinary look is one I wouldn't have thought to do, but really does explain a lot. I'm torn, because I'm sympathetic to most of the defenses. It is hard to really believe we are making meaningful progress, though. Even if I am enjoying many of the small practical improvements we have managed to get out of things.

Erlich_Bachman 8y ago
What would "meaningful progress" entail for you? How would your life change, how would the life on earth in general change? Great achievements in science (aided by AI)?
- taeric 8y ago
  That is a different question, though. The question if if there is meaningful progress in ai. The amount of progress explained solely by percentage improvements against a benchmark is pretty high. I don't think this is particularly damning. Nor do I think it should be halted. However, I agree it is hard to call progress.
  - Erlich_Bachman 8y ago
    What specifically would be easy for you to call progress?
    - taeric 8y ago
      An equation that could predict a better ML model. Honestly, I think claiming I want an equation is a touch too much. However, a falsifiable prediction would be nice. Imagine if the only way we got accurate ballistics was by requiring faster more powerful guns all of the time. "We can hit the target, but only if we upgrade our guns to railguns and limit ourselves to large targets."

jmickey 8y ago

Relevant - AI Progress Measurement page by the EFF: https://www.eff.org/ai/metrics https://www.eff.org/ai/metrics

toolslive 8y ago

For typical problems (playing chess, tissue segmentation, translation, ...) you have the quality of the solution (sometimes difficult to measure) versus the cost of achieving it (energy/enthropy and maybe time). The cost is important.

stared 8y ago

Please, no. While he provides some food for thought, asks a few important questions (and provokes even more), this text is full of pseudointellectualism. As in: fancy words but utter lack of understanding of the core subjects one writes about. Most importantly: what is science (no, it is not a trivial question; I recommend going Ludwik Fleck's "Genesis and development of a scientific fact" route, http://www.evolocus.com/Textbooks/Fleck1979.pdf http://www.evolocus.com/Textbooks/Fleck1979.pdf) and what is AI (it is a vast field; some of it IS math/CS (as in: proving things), but it is a small part). For example, many of things he says would fit other practical disciplines, e.g. medicine. Yes, experiments are on a group of people. Yes, criterion that "drug X works better than drug Y, but we don't know why" is sufficient. > I don’t know data science folks well, but my impression is that they find the inexplicability and unreliability of AI methods frustrating. Well, it is not the main issue (speaking as a data scientist working with AI). At least he acknowledges his lack of expertise. > These failures of scientific practice seem as common in AI research now as they were in social psychology a decade ago. From psychology’s experience, we should expect that many supposed AI results are scientifically false. Also - no, it is not at level of psychology when it comes to the replication crisis. A lot of code (though, unfortunately, not all) is shared online, by the authors or other contributors, and people do replicate it (or there is an absence of replication, which also conveys a message).

Radim 8y ago
Speaking as both a ML researcher and an applied ML business owner (one who hired stared at one point — hi Piotr :), I respectfully disagree. The replication crisis in "AI" may not be as bad as psychology (I wouldn't know), but it's not great. Sadly, my brain has somehow learned to equate "SOTA" with "hot-stitched crap, stay away". Too many painful lessons. On the subject of publishing code: this is useful to the degree that it removes bad faith as the possible reason for the lack of replicability. But otherwise helps little in practical terms. You just have the privilege to sift through the bugs and bad design in more close-up. "I am afraid you are right. I used to reach ~72% via the given random seed on an old version of pytorch, but now with the new version of pytorch, I wasn't able to reproduce the result. My personal opinion is that the model is neither deep or sophisticated, and usually for such kind of model, tuning hyper parameters will change the results a lot (although I don't think it's worthy to invest time tweaking an unstable model structure)." = quote [1] for one of the "new SOTA" papers from NLP (WikiQA question answering), where the replication scores came out 62% instead of claimed 72%. I generally call this the "AI Mummy Effect" — looks great but crumbles to dust on touch. [1] https://github.com/pcgreat/SeqMatchSeq/issues/1 https://github.com/pcgreat/SeqMatchSeq/issues/1

sgt101 8y ago

If the subject is seen as a system rather than a thing you can make some progress. On the one hand we have Artificial Intelligence, where insights from cognition and biology are used to model and explain reasoning and behaviour. On the other we have AI where people use the outcome of Artificial Intelligence research with other pragmatically selected technical components to develop technology. I think that there is interdependence and exchange between the two, but they have different methodologies and processes. I also think that huge trouble is created when members of one community use the other communities clothes and achievements. For example Artificial Intelligence researchers talking about the practical impact of AI and near term application of their work and conversely AI researchers claiming that their work "works like a brain".

jfjfjfjfyy 8y ago

When AI progresses, it evaluates you, not the other way round. By the time this is realised, it is already too late for yet another 'generation' of AI researchers.

laichzeit0 8y ago

> “This year, we’re getting Z% correct, whereas last year we could only get (Z-ε)%” does sound like progress. But is it meaningful? This is one thing that bother's me a lot when I read published work. I have a feeling this is a result of everyone using the same benchmark datasets, so it inevitably becomes more of an _engineering_ exercise rather than scientific progress. In NLP the difference between a publishable result and one that is not is often to squeeze out a few extra (Z-ε)% by throwing in an attention mechanism and ensembles to your new super duper improved SOTA architecture. This is the problem of "replicability", which really requires more than just the same benchmark dataset used over and over again. The author seems to touch on this point later on. Then there's the issue of "reproducibility". Very few researchers seem to publish their code with instructions how to build and re-create their results. What an awful lot of time is wasted trying to reproduce results. Here's a good example: https://groups.google.com/forum/#!topic/word2vec-toolkit/Q49FIrNOQRo https://groups.google.com/forum/#!topic/word2vec-toolkit/Q49...

ggm 8y ago

Most (human) languages have a set of rules and a variant set of variances ("i before e except after c, and this catalog of things which we pretend don't exist") so ML and NLP are cases of things which are ameanable to rules based systems, and because english corpus exists widely can be tested openly, against each other and a common norm of comprehension (for english speakers) generalized AI does not lie here: systems which uncover the grammer rules and exceptions do not generalize to systems which uncover rules in Law, or Equity, or financial trading, or other things. Yes, you can train nets. But the commonality here, is you can train, not that emergent AI is found. (not an AI person, strongly anti-AI perspective from life experience in compsci)

fny 8y ago
Is there such a thing as AGI really though instead of collections of modules that are trained and may have no meaningful mode of interaction? What does NLP have to do with the mathematics of financial trading? AGI has always felt to me as a large scale interdisciplinary/intercomputational activity, much in the same way that most human intelligence derived from years of intergenerational and interpersonal intellectual development. AGI will never be one but many: many intelligences and systems interacting to produce something of utlitarian value.
- ggm 8y ago
  much in the same way that most human intelligence derived from years of intergenerational and interpersonal intellectual development. wow! I had no idea the archeology on early hominids was that good... (yes, implicit /s)
taeric 8y ago
This treats human language as a somewhat static thing that can be fully specified. Which, interestingly, avoids a large portion of human languages in poetry and metaphor. A bit like judging story telling by only accounting true stories, thus completely ignoring all of fiction. Saying nothing of false stories purported as true. It is quite possible that understanding is nothing more than a statistical relating of otherwise unrelated things. Some things, we accept as being 100% related. But, quite a lot of things are just "highly related." Where "highly" is probably not specified.

YeGoblynQueenne 8y ago

>> GOFAI had several defects, but… the main thing is, nearly all of it was false. It would be nice to see some kind of substantial examples of how "(nearly all of) GOFAI was false", accompanying statements like the one above. The problem of course is - those are very hard to come by. That is so because logic-based AI was abandoned. And it was abandoned because funding was cut repeatedly, not because of its failure to prove this theory or achieve that aim, but because the ones holding the purse strings were administrators and military pencil-pushers, who had no way to know a successful, or failed, program if one came up and bit them in the boogies. And just to substantiate my comment- what, exactly, was "false" about logic programming, one of the major research subjects in GOFAI? It worked just fine back then, it works just fine right now. In very practical, down to earth terms, you can prove a proposition, or a predicate, true or false by automatic means, sure as you can answer "2 + 2 = ?". So, really- more substance, less assertiveness, would do a world of good to those for whom "AI" means everything that they read online after 2012 and who may end up missing a hell of a lot of the history of the field if they take that sort of "GOFAI failed" statements at face value.

artilect 8y ago
Aigo.AI has spent a few million attempting to build AGI with a GOFAI architecture
colorint 8y ago
As a general matter, you can't prove a predicate true or false by automatic means. Logic programming is just as "artful" as imperative or functional programming, because they all run into the same problem: it's impossible to tell the difference between long-running and infinite computation. The question of algorithms for general logic was explicitly addressed as the Entscheidungsproblem, meaning decision problem, which was independently proven undecidable by both Turing and Church: https://en.wikipedia.org/wiki/Entscheidungsproblem https://en.wikipedia.org/wiki/Entscheidungsproblem
- YeGoblynQueenne 8y ago
  >> As a general matter, you can't prove a predicate true or false by automatic means. Not in the general case, sure, yet in practice I'm sure we've all written plenty of code that terminates just fine. [Edit: I'm talking about imperative as well as logic or functional programming code]. The question then is- what does it mean when a program terminates? In the case of principled approaches like logic or functional programming, you have a pretty good idea what that means (e.g. a logic program proves a theory true or false). When an imperative program terminates, it's a very hairy affair to say what, exactly, termination means. [Edit 2: Actually, if you think about it, there's nothing we can really achieve in the general case (including machine learning; see language learning in the limit). In practice, on the other hand, we're doing things, alright - by continuously relaxing principles and fudging limits as necessary (see PAC learning)].