TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

What intelligent machines need to incorporate from the neocortex

153 点作者 dendisuhubdy大约 8 年前

12 条评论

return0大约 8 年前
Recently Matt Taylor from numenta had a reddit AMA (<a href="https:&#x2F;&#x2F;www.reddit.com&#x2F;r&#x2F;artificial&#x2F;comments&#x2F;6beeqj&#x2F;5182017_1200_pm_pst_iama_with_matt_taylor_numenta&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.reddit.com&#x2F;r&#x2F;artificial&#x2F;comments&#x2F;6beeqj&#x2F;5182017_...</a>). I was disappointed to find out that numenta does not really have collaborations with neuroscientists to test or adapt their theories. Their theories in general are not well known to computational neuroscientists either. In that sense, i m not even sure about the authority of numenta on neocortical theories.<p>For example, the article mentions the ability of clustered synapses to act independently, but , on the one hand, it has been shown independent dendrites can be approximated as an extra neural network layer (so they ARE covered by today&#x27;s ANN approximation) , and OTOH there s a number of papers showing that synaptic clustering does not exist in sensory areas. And learning by rewiring is basically the introduction of random connections which persist only if their weight increases enough (roughly corresponds the continuous formation of filopodia and the fact that large spines persist longer).<p>Machine learning at the moment is an empirical science that has made great strides without consulting neuroscience for it. I think that has been a good thing: without having to bend towars some biological plausibility researchers have been more exploratory and creative, which has led to the creation of an empirical body of knowledge from which neuroscience could benefit in the future. OTOH, having watched the field of computational neuroscience there has not been a lot of progress since , basically the 80s. So i believe it would be best to leave each of the two fields go their own way.
评论 #14474403 未加载
评论 #14474870 未加载
评论 #14493151 未加载
评论 #14475194 未加载
computerex大约 8 年前
For those interested, Jeff Hawkins wrote a book called On Intelligence and he is largely the guy behind: <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Hierarchical_temporal_memory" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Hierarchical_temporal_memory</a><p>Deep learning proponents focus on solving specific problems using mathematical models inspired by biology, whereas HTM proponents argue that biology is important and it should play a more central role. Deep learning folks are more geared towards applied AI, whereas HTM&#x27;s are way more ambitious and are trying to solve the problem of intelligence&#x2F;AGI.<p>Numenta, the company behind HTM&#x27;s has a platform called NuPic for &quot;intelligent computing&quot;: <a href="https:&#x2F;&#x2F;github.com&#x2F;numenta&#x2F;nupic" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;numenta&#x2F;nupic</a><p>But as far as I know HTM&#x27;s have never been applied to anything non-trivial successfully.
评论 #14473517 未加载
评论 #14474664 未加载
amelius大约 8 年前
And perhaps what the neocortex needs to incorporate from intelligent machines: backpropagation?<p>I mean, so far we have no biological evidence of backpropagation, and it seems pretty useful.
评论 #14475466 未加载
bobosha大约 8 年前
I have followed Hawkin&#x27;s theory from its inception ca. 2005, since then they have been through multiple evolutions: from hierarchical bayesian models (with then collaborator Dileep George) and now an Sparse representation with Subutai. However, they have struggled with a &quot;killer app&quot;. Barring some toy examples, very little by way of real-world use-cases.
shahbaby大约 8 年前
Strongly believe that if anyone is going to solve human level intelligence in our lifetime, it&#x27;s going to be Numenta.<p>&#x27;But we can&#x27;t even figure out worms!&#x27; - Worms are made up of neurons but they perform a different function from what the neocortex does so studying them is not like studying a simpler problem, it&#x27;s studying an entirely different one.<p>&#x27;But ML can do X better!&#x27; - Unlike industry or academia, Numenta&#x27;s primary goal is to figure out how the neocortex works, it&#x27;s not about profit or publications.<p>&#x27;Biology contains details we don&#x27;t need!&#x27; - Numenta&#x27;s approach is not biologically inspired, it&#x27;s more like biologically constrained. They avoid implementations that are <i>functionally</i> different from how the neocortex works.<p>I would highly suggest reading On Intelligence to learn more.
评论 #14474927 未加载
评论 #14482453 未加载
danharaj大约 8 年前
How would one go about self-studying neuroscience to the level of say, a second year grad student not yet specializing in anything? That is to say, familiar with the basic concepts and able to make sense of current research given enough time.
评论 #14474340 未加载
评论 #14473828 未加载
评论 #14473816 未加载
评论 #14475482 未加载
SAI_Peregrinus大约 8 年前
&quot;Machines won’t become intelligent unless they incorporate certain features of the human brain.&quot;<p>How arrogant! The features they list aren&#x27;t unique to humans, or even to mammals. They&#x27;re all present in the nidopallium of birds as well. Machines won&#x27;t become intelligent unless they incorporate certain features of intelligent life.
meri_dian大约 8 年前
&gt;&quot;Intelligent machines don’t have to model all the complexity of biological neurons, but the capabilities enabled by dendrites and learning by rewiring are essential. These capabilities will need to be in future AI systems.&quot;<p>This is the fundamental challenge that we face. Our ability to build AI that can emulate human thought is not limited by a poor understanding of the brain. Our current theoretical understanding of how human cognition arises from neural processes is probably close to a level sufficient to build human level AI. What limits our progress is the staggering computational demand of simulating a massive network of highly dynamic units.<p>Shortcuts, simplifications and clever algorithms will only get us so far. At this point, processing power rather than theoretical understanding is the limiting factor.
评论 #14473634 未加载
评论 #14473491 未加载
评论 #14473766 未加载
评论 #14473463 未加载
评论 #14473856 未加载
andbberger大约 8 年前
numenta makes my kook sense tingle... but Jeff Hawkins did found the Redwood Center for Theoretical Neuroscience (now part of HWNI), and for that I am grateful.
评论 #14476281 未加载
mr_overalls大约 8 年前
&quot;The question of whether Machines Can Think... is about as relevant as the question of whether Submarines Can Swim.&quot; --Dijkstra (1984) The threats to computing science (EWD898).
theprop大约 8 年前
Some professor asks students on the first day of neuroscience, if all there is to know about the brain is a mile long, how far have we gotten so far?<p>Students guess a few yards, a hundred feet, ten feet...the professor says no no, not even 3 inches.
gene-h大约 8 年前
&gt;Our discovery is that every region of the neocortex learns 3D models of objects much like a CAD program. Citation needed. If this was the case, certainly we&#x27;d have some experiments that show this? Especially since every part of the neocortex is supposed to be doing it.<p>Now if we can make artificial neural networks that work with 3D data, learning such things as 3D data to value, 3D data to 3D data mappings that would be damn useful. IE, estimating how much it would cost to make something from a CAD model or how aerodynamic a thing is without running costly CFD.<p>I&#x27;d also argue that we don&#x27;t need &#x27;truly intelligent machines&#x27; to &quot;build structures, mine resources, and independently solve complex problems&quot;<p>Ants and termites are capable of doing similar tasks and I&#x27;m doubtful the author considers them &#x27;truly intelligent&#x27;.<p>&gt;it should be possible to design intelligent machines that sense and act at the molecular scale. &gt;These machines would think about protein folding and gene expression in the same way you and I think about computers and staplers. &gt;They could think and act a million times as fast as a human. So if the author means in simulated environments, we are quite slow at simulating molecules. For molecular simulation, we need something like femtosecond(10^-15 s) time steps whereas each time step is on the order of milliseconds. We are trillions of times slower than realtime. This is for completely classical systems, if we take into account quantum effects, it&#x27;s much longer. Oh and our simulation methods for such things are terrible. Intelligence would help here, but it&#x27;s not going to be millions of times faster than a human.<p>Now if they mean videoing what&#x27;s happening with a microscope and learning from that, well the problem is we don&#x27;t have a perfect microscope for seeing such things at the nanoscale. So in order to figure out what&#x27;s going on we have to get creative and make tests for each thing we&#x27;re trying to analyze.<p>Now if the author means nanorobots inside cells doing learning and what not, just having such machines would be useful in and of itself. Heck if we could make such things, we wouldn&#x27;t need to worry about problems such as gene expression or protein folding because we&#x27;d be able to make our own damn proteins or our own damn cells for that matter. Even with drexlerian tech doing this sort of machine learning at this scale is pretty ridiculous. Current nanobot designs require something on the order of kilobytes of memory[1]. In addition, gene expression, protein synthesis and folding are slow processes. Average protein synthesis time for eukaryotes is 2 minutes[1](eons as far as simulating these things is concerned!). So getting this data can&#x27;t happen much faster than a human can think.<p>[0]<a href="http:&#x2F;&#x2F;people.umass.edu&#x2F;bioch623&#x2F;623&#x2F;Second.Section&#x2F;7.%20CoT.Web.2.08.pdf" rel="nofollow">http:&#x2F;&#x2F;people.umass.edu&#x2F;bioch623&#x2F;623&#x2F;Second.Section&#x2F;7.%20CoT...</a> [1]<a href="http:&#x2F;&#x2F;www.rfreitas.com&#x2F;Nano&#x2F;Microbivores.htm" rel="nofollow">http:&#x2F;&#x2F;www.rfreitas.com&#x2F;Nano&#x2F;Microbivores.htm</a>