@gmirey dropped a nice link on this earlier:
4 Likes
Update: almost 2000 users.
Update: almost 150K views total
9 Likes
By comparison:
Crochet Guild of America - 3,000 members
Always Elvis Fan Club - 16,000 members
9/11 Truthers - 25,000 members
Looks like youâre behind.
1 Like
I think a cellular automata like network but using dynamic connectivity, could probably exhibit rhythms and other brain like properties, perhaps even general intelligence, while potentially being much simpler in mechanisms.
I am new to this forum, I am a bioinformatician and heavy user of Deep learningâŚ
To me, Numenta approach is much more interesting and profound than deep-learning and other ML methods, but the ML community is more interested in what works. Deep learning works in many cases ABOVE the average human performance, and hence, it is useful in many areas. But eventually it is just a powerful function-approximator.
Nevertheless, I have got the feeling that the ML field is moving toward Numenta direction faster than we think. Late research showed a significant improvement in unsupervised / self-supervised approaches which is the real "cake" of ML as Yan Lecun stated.
In 2018 we learned that unbelievable results can be achieved with system that just try to âmodelâ the word, like: self-attention, language models and generative models. Thus, even in the deep-learning world, it is becoming clear that the future paradigm of ML is a not a data-angry classifier, but a strong model of the world, learned from sequential or forward predicting model (like GAN), which shockingly similar to Hawkins view of the brain: A âconstantly changing multiple predictive models of the worldâ â In my opinion it is not far from what the recent âBERTâ and âStyleGANâ models are, although the architecture is very different.
The amazing thing about the brain is how it can learn such a complex model of the world without much data and with a very low-powered and slow hardware. This is where Numenta should focus: The future AI technology is probably not a data-angry monster running on multiple GPUs, but a context-aware, self-supervised, ensemble of learners, that can learn numerous concepts and objects, and have a âcommon senseâ about the world like human does. HTM has a much better chance to go there than DL.
I think that Deep learning is here to stay, but it is just a step toward a brain-like system of the future. We just need to go through GANs, Attention, Capsules, Language-models, etc. before hitting HTM again, probably with some clever take on current computation paradigmâŚ
If the Numenta software will be as mature and elegant as keras or pytorch and will show some interesting unsupervised results on real problem, then it will fly. I donât think that it has a chance to beat DL in big-data supervised tasks, where humans canât.
8 Likes
Thanks for your thoughts. I agree with a lot of what you said. I think HTM will require input from multiple sensors in space to really shine on hard tasks like object modeling. We need to split up the senses so we have different cortical columns modeling different locations in space, attached to different sensors. That is how sharing lateral connections will enable sensor fusion. We think this is how local patches of one sense are fused, and we also think this will apply across sensor boundaries.
But one thing we can do with todayâs tested theory is see how it can be applied to the Deep Learning systems of today to improve performance or accuracy. That is currently what we are looking into, and I have to say that from what Iâve seen so far I am encouraged that we can make some impact on the current Weak AI scene by adding some HTM ideas.
4 Likes
Hi everyone. Iâve been reluctant to join the conversation for a couple of years, for many reasons, and will restrain myself further, as writing posts like this one takes too much time. But here are my .5 on a given topic.
TL;DR: to succeed, the HTM ecosystem needs systems, processes, practices and tools that belong to general software development domain, or that are spread across multiple domains: system programming, network programming, game programming, to name a few.
First of all, yes: these algorithms sure are non-trivial, if not hard. But implementing one is totally feasible nonetheless. As with any software, developing and polishing an algorithm is an iterative process (even/especially when itâs backed with massive amount of scientific data). Hence, it requires a feedback loop.
An effectiveness of such feedback loop depends on a development process. How quickly you can get the results from a working version of the algorithm or an integrated system? How quickly you can decide whether the system behaves correctly and does what it meant to do? Are you sure you get clear results? Etc.
These are questions that are being asked every day in general software development domain. The libraries, frameworks and tools are used to make the process easier and more complex tasks to be more âtackleableâ (is there a more appropriate word for this? Iâm pretty sure there is).
From what Iâve seen, HTM ecosystem lacks that feedback loop almost completely. This is a complete show stopper.
The second one. These statistics algorithms, or ML/DL they called these days - you donât need them to succeed. And they wonât get you, the theory, and the HTM algorithms anywhere. They might bring some tools and practices into development processes though, but not much.
More than anything else, you need a system that scales; and scales very well, and doesnât consume all available resources while constantly operating. This architecture needs to be the foundation of the solution, and you need to build it first. What kind of task the solution will perform is insignificant, because any algorithm can be embedded in it.
And then, you can elaborate processes and feedback loops on top of that system. And improve it further, and improve algorithms you embed.
I was pretty surprised when I saw the news that Numenta is hiring ML/DL developer (and looks like the position is already filled since the announcement is gone). Numenta, and/or the community, needs a couple of good software generalists, with strong production-grade experience in network programming at least. They might not be familiar with some concepts, but they can build an infrastructure that you can use. Scientists donât construct laboratories; instead, they formulate requirements to be met.
I wonât discuss the theory itself here, since it doesnât matter much; no one have been able to prove or refute it because there is no clear process (of proving/disproving) to follow, yet. Whether the theory is right or wrong (and I assume itâs right, at least in its core principles), a well-defined process, that is easy to follow, will get you to the point more quickly.
4 Likes
Jeff: âI think we are pretty mainstream, we are just rare.â (source)
@rhyolight canât wait for Episode 16 to come out. 
3 Likes
While I donât have a binary opinion on this, Iâm more inclined to this opinion. Going DNN or integrating DNNâs might lead HTM to a trap. One of the things I liked about HTM is that even though the algorithms are complex and involve self-organization and probablistic techniques, they are still mechanical and can be modelled with classical CS computational models. IMO this is something that is extremely important both for engineers and businesses so that a certain level of confidence can be established about how an HTM application will operate in the real-world just like any other non-AI software out there.
2 Likes
Donât worry we are not going Bayesian. Weâre just trying to see what tricks we can pull off in the current ML space to garner some attention. Maybe there is some low-hanging fruit where HTM ideas bring big gains? It is a good time to ask these questions.
4 Likes
HTM isnât mainstream because there isnât any public relateable news about it. Deep Neural Nets went mainstream the second they defeated world grandmasters in Go. Funny thing, I wanted to try making a Go AI using HTM before DeepMind existed. At the time I didnât have the skills and there wasnât a Windows version. Anyway, people relate to games. If HTM can play games, people will pick it up, especially if you can pit it against DeepMind. People would also be more likely to pick it up if itâs just the download of a library.
1 Like
I agree with much of what has been said in the discussion about why HTM isnât mainsteam yet.
I think that it may be necessary to depart somewhat more from biology to make it more useful in the short term. The lack of supervised learning / reinforcement learning is quite a problem, in my opinion, in terms of applying it to most ML problems. What I imagine may bridge the gap is a combination of HTM with another deep learning approach.
Imagine the following: HTM processing of image data -> sparse representation output -> convolutional neural network -> object classification.
To be useful for most problems, it needs what I would call a decider network. It needs a way to choose between various responses and then be reinforced based on accuracy. Youâll never have AGI without this IMHO. Humans and animals learn through an immensely complex process of supervised and unsupervised learning. Humans in particular have very sophisticated decision, reinforcement, and punishment systems.
Has anyone tried combining HTM output with a supervised learning approach?
Best regards,
Jack
I am planning to implement something opposite to thatâŚ
(1) Deep-learning as the âsensorâ to extract image features (ResNet trained on ImageNet)
(2) encode dense space of the DL-sensor last layer into SDR using some clustering approaches.
(3) using multiple HTM columns to learn this representation in an âunsupervisedâ way. (I was thinking of simulating âmovementâ of the detector over an image as a sequence of data).
(4) using this HTM-ensemble to do âzero-shotâ object-classification / object detection.
The idea here: use DL in supervised-learned pattern recognition, and use HTM where it shines: making sense of the world and model the world in an unsupervised way.
The test case here will be to perform âzero-shotâ object recognition/detection âŚ
given an UNSEEN object class: can an ensamble of HTMs recognize new categories?
Can it tell for example that some unique animal it hasnât seen before, is somewhat similar to animals that it has seen?
3 Likes
There are demos from Cortical.io doing just this with words of animals (inferring the eating preferences of Coyotes after seeing those of Wolves and Dogs), though I donât know of anything like that with imagesâŚ
1 Like
That sounds like an interesting idea. Iâm just not certain how you end up with a decision or meaningful output in this case, but it sounds like you have ideas for that.
I feel like youâd still need a softmax layer on top of the HTM stuff to train as a classifier and help determine the error (which would trigger the HTM system to engage in itâs hebbian learning process). Afterward, when youâre happy with the performance of the classifier, you could either ditch the softmax, or keep it in parallel with the HTM system.
Thatâs where Iâd start.
1 Like
I came across this thread through some google research and a few months ago was the first time i read about Numentas approach. I always felt something is missing in the approach of deeplearning as it just neglects some very intuitive details about the brain should operate and my gut feeling is that an approach like Numenta could potentially cover a lot more ground to do âbrain workâ than the deep learning approach.
And i guess it has been posted here before, but i found this remark by Hinton in that regard quite interesting: Hinton comment on Backprop
I found it esp. remarkable since he always was a firm defender of this method I guess its a scientist´s trade to stick with a method they really believe in for a long time until there comes a monumental event that makes them change their mind or they go âout of businessâ.
That being said, i do like what Numenta is producing and i will start following this more.
6 Likes
A short comment/suggestion on step 2 @Dror_Hilman - maybe instead of using a clustering approach, you can add a regularization term to your loss function to force weights in the last hidden layer towards 0 or 1, and clip them afterward to binary. You can also enforce the amount of sparsity you are expecting by using some sort of dynamic pruning. Of course, this approach means you would have to fine-tune the feature extractor as opposed to just using a pre-trained model with no extra learning.
I would be very interested to see your ideas implemented, please share in case you go forward with it.
2 Likes