I figured I might as well through in my introduction. Why not. I’m an AI researcher from America but I have been working between Spain and the UK for the last two years. I’ve been programming since a little bit before 2008 but I really got into it in 2009 when the first free version of Unity 3d came out. I’ve been programming ever since.
I have only been in AI for a little over 2 years, but I quickly grew tired of it because I had fanciful ideas that it was something related to real human intelligence. I started digging into what makes biological intelligence so different than the statistical approach that traditional AI has taken. I started with the simplest form of intelligence first, ameobas. With no brain cells or mind, they are capable of tracking temporal changes in nutrience and environmental changes as an individual. In a group they are capable of much more.
I started looking into seeing if anyone had made “ameobic intelligence” like models. Some of the first results around many iterations of search queries I was making was from Numenta. It’s the closest thing I have seen to true biological intelligence that exihibits all of the main qualities. It got me excited of the possibilities around digital intelligence again and I have been digging into it for the past month or so in private outside of work.
I definitely think there is a future to it and I am excited to get working with it or at least some concept based on the findings and research around it. So excited I’ve actually finally quit my job here in Spain and I’m headed back to the US to start my own private research and likely start a company up with a freind. I really do think it is the best shot towards general digital intelligence.
Definitely looking forward to contributing pieces to the edge of human knowledge.
Hi, I am from Assam(INDIA) and I want to know myself better. I discover Numenta when I was searching about “new intelligence model” in this year(2018) and somehow it was linked with “cortical.io”. I don’t work in anywhere, I’m still studying in home.
I was so much fascinated about something when I was 12 years old and later I found out, that something is AI. I am not satisfying with my life since still It’s not happening what I am trying to.
Anyway, I am still trying.
I salute you all.
Hi, my name is Gien. I’ve been wanting to join the forum for awhile but found it too cumbersome in the past so I’m glad you guys converted over to a more standard web forum format!
I’m here to learn if we can take Numenta ideas and apply to a system architecture we are building for a festival we are launching in late 2019 called The Tipping Point Festival. Our tagline is: Reaching Social Tipping Points before Planetary ones.
The Tipping Point Festival (TPF) is envisioned as an umbrella framework to mobilize a Bottom-up movement to accelerate rapid transition to a human civilization that is living within the limits of nature.It’s synchronized with the Mission 2020 objectives. Our feeling is that it will take more than governments and big business to avoid tipping points. The commons, the billions of citizens around the planet is the missing element, the sleeping giant of change, but it needs to be organized in a systematic way. TPF is designed to provide a space to catalog all the change actors and projects, networking them, and using a combination of Human and Artificial Intelligence (HAAI) working in tandem to solve the growing set of problems that threaten civilizational stability.
I’ve been interested in Jeff’s ideas and work for years and glad to see him and his team making such great progress. I especially like the fact that Numenta takes a biologically inspired AI approach. We are interested to see if we can use these ideas to develop a framework for both learning existing ideas and encoding new ones. We are big on open source and our platform is a problem solving space to create new innovation that goes directly into the commons because we feel that if billions of people are going to be empowered and not get left behind, they need the greatest accessibility possible to new innovation. For this reason, we are currently exploring creating an Open Knowledge Commons Foundation to help fund and systematize open knowledge. Numenta offers a possible practical solution for us to develop such a framework.
One question I wanted to ask is how Numenta and Redwood’s research synthesizes with the recent results of the Blue Brain project, specifically the new discovery of higher dimensional structures/geometries that Henry Markham believes can store memory. https://bluebrain.epfl.ch/cms/lang/en/pid/146326
other interesting research news that could link to Redwood/Numenta https://bluebrain.epfl.ch/page-141408-en.html
Hi, I’m Felix. I’ve enjoyed reading the threads here and thought it would be good to get an account so that I could track what I have/n’t read. Keep up the good work.
My name is John Grisinger. I’m a civil engineer by training whose career soon moved into information management and database architecture where I did some consulting. I wanted to understand what information is and that required understanding what memory is and how we acquire it. I wondered if there might be a way to use business rules to directly build business systems. That led me to study linguistics to better understand the structure of languages. I read On Intelligence when it first came out as well as a variety of other books and papers. There was little that was helpful on my specific interest. I looked instead for the simplest possible platform that allowed all forms of information to be stored as well as to store the elements of language. What I developed was essentially a directed graph model, but one that is simpler and also more general than the graph models currently implemented. I demonstrated that it could store all of the forms of human comprehension about the world and the word needed for language. I identified how it could generate both linguistic and mathematical expressions from what is stored. That platform looked like how neurons might be configured. I used reverse engineering to find neural behaviors that store neural configurations that replicated the configurations that could be stored in that platform. The neural behavior I arrived at is surprisingly effective at explaining how perception and memory arise from neural behavior, how recognizing and recalling work and provides a more complete description of the different forms of memory.
I am writing a paper describing that memory-specific neural behavior and the configurations of neurons it creates. I currently refer to it as Perception Pattern Structure Theory. I have completed a draft of the first part discussing neural behaviors, perception, stored patterns plus recognizing and recalling what is stored. I was attracted to Numenta as a forum for this paper because of its research interest in the cortex, its openness to new ideas and its recognition of reverse engineering as a tool to better understand cognition. Since my interests are somewhat divergent from HTM, they appear to complement those I have seen in the forum’s members. I plan to post that paper shortly under Tangential Theories.
This is a very remarkable group of people. Most of the group–I am guessing–are scientists and researchers. I am coming from outside this hard-core perspective, so I hope that my comments will be welcome. My doctorate was in optometry and my post-graduate work was in Orientation and Mobility (Blind Rehabilitation). Essentially, after 33 years of teaching children in special education who had navigational disabilities, I became quite knowledgeable about how human beings navigate. I am now retired and writing books about the evolution of consciousness and about a systems theory that I call the navigational-brain. This navigational mind/ brain/ consciousness is a global perspective–an attempt to synthesize knowledge.
I found this community at Numenta when I was Googling the hippocampal set of cells that are so navigationally-specific. I think I know why these cells evolved and how they relate to a broader theory of mind. I was just trying to send a message directly to Jeff (his book On Intelligence greatly influenced my early thinking), but I don’t see a way to do that. I feel like my systems theory is very timely given the interest in grid cells as relational systems that model the way we socialize and think. Again, I think I know why that would be.
I put together an 8 page summary of the navigational-mind theory for Jeff and just wanted to give it to him. I am sorry that I don’t know the other brilliant people who are working with him–it doesn’t matter who gets this knowledge. I just want to give it away. It is frustrating to have this body of knowledge but have no one to discuss it with! Anyway, I am grateful that this forum exists and I look forward to a dialogue.
Thanks for your suggestion. I would certainly appreciate dialogue around the theory of the navigational mind. I will see if I can start such a thread. Do you know why my summary of the theory–a microsoft doc–was blocked as unauthorized?
Sorry about that, sometimes the spam filter is aggressive, when it saw you trying to post a document to the site it said no thank you. I’ll help you out with this offline.
I’m Chris, a 30 year vet of enterprise software with a longer history as a brain-fan kicked off by Carl Sagan’s “Dragons of Eden” in the 80s. I’m a federated, distributed processing/transactions, big data, high performance SQL/Java/Python guy that started long ago with HPUX, DGUX, and 10 years of NeXTSTEP.
I spent a few years on topology and fiber bundles - which has come in handy in recent times for topological data analysis / data science.
I’ve gone through spurts of obsession with language, cetacean (woot Orcas!) intelligence, ravens, and more recently octopus. “The Other Brain” pulled me into glial cells. Word2Vec pulled me into embeddings, auto encoders, eigen decompositions, the rotation problem and tensors - blew my mind that abstraction == compression. DeepMind’s “Hybrid computing using a neural network with dynamic external memory” pulled me into the perforant path, index theory of memory, and stochastic associative content addressable memory. Currently fascinated by Santoro et al’s “relation networks” - how does reasoning work? (Also looking forward to attending DARPA’s AI conference with heavy reasoning focus in March if you’re going - say hi!)
Then found HTM-School on Youtube talking about sparse distributed representations, which pulled me into Pentti Kanerva’s book and to HTMs and Numenta.
All of this is a blast! It’s some of the most exciting stuff I’ve ever encountered and the weird introspective a-ha! moments are too much fun.
Yes @rhyolight - I’m the fovea-question guy. Thoroughly enjoyed the hangout - could listen to that style of discussion for hours a day! Thanks for putting it together.
@rhyolight: outstanding. Now, if only keeping busy were the problem. #Need48HoursInADay Looking forward to going through your “podcast with a neuroscientist” too. Glad Numenta pays you to do what you do and glad you’re doing it.
Hello! I have an undergraduate degree in philosophy, and I have a decade-long love for philosophy of mind and cognitive science. Lately I have been getting into AI and machine learning, which started with my obsession with autonomous cars. I even started a forum to talk about autonomous car technology because I couldn’t find a good one anywhere online.
I actually have no idea what direction I want to go in, in terms of career or education. I would like to do something related to theorizing about brains, minds, intelligence, consciousness, and cognition, but I have no idea where I fit into the world. I feel like I have a strong knack for philosophy, but I think philosophy of mind has to be combined with cognitive science to be interesting, or to make progress.
What I love most, and what I’m best at, is theory and theorizing. I would love to dedicate my life to theoretical cognitive science. But I don’t know where I could go to do that. I don’t know if theoretical cognitive science even really exists as a field yet. When I say “theoretical cognitive science”, I mean like what Daniel Dennett does in Consciousness Explained. He creates a high-level theory of human consciousness, while knowing enough about psychology, neuroscience, and AI to give a rough sketch of the implementation details. Somewhere between armchair philosophy and full-on science or engineering there is theoretical cognitive science.
I find Numenta’s work interesting because they’re developing high-level theories of cognition, although in Numenta’s case they’re also doing experimental science and apparently neural network engineering.
I suggest you consider a career that includes writing about the topics you find interesting. This is about as close as you are going to get to be paid for an interest in a topic without being a worker in the field or a teacher.
Not sure what you mean, “theory” is in the eye of the beholder. But I think mine is as high-level as is gets, short of becoming meaningless: http://www.cognitivealgorithm.info
I’m going to be a little bit different and add some structure.
Name: Jeff Hykin Location: Texas A&M University Occupation: Senior undergrad in Computer Science, with a previous major in Cognitive Science Claim to Fame: If you write C++ in Visual Studio Code, you’re using one of my repos Career Plan: I will probably end up in the Bay Area since I’d like to get a masters and PhD in AI, and work at a company similar to Open AI, Vicarious AI, Curious AI, or Numenta.
Reasons for being on the form:
I care very little for non-biologically based machine learning, e.g. self driving cars, factory optimization, etc. This community is one of the few places I can find where AGI discussions take priority over domain-specific optimizations and applications.
I aim to work in the balance between neurological research and computer science implementation of true AGI. And Numenta seems to be spearheading the neurology side.
I read On Intelligence in high school and again during sophomore year of college. It was one of the most clarifying books on AGI that I’ve read.
Like many of you, my ultimate goal is to help bring about truly human-like AGI.
Relevant AGI Skills:
Sadly, I have not yet dived deeply into the HTM School. I’m trying to learn it slowly it on the side.
Took Andrew Ng’s Coursera machine learning course 4 years ago.
Read most of the AI books available on Audible.
I more or less keep up with machine learning methods: ConvNets, GANs, Q-learning, and Capsule Networks.
Not much experience with machine learning libraries (Tensorflow, pytorch, keras). This is my next major focus.
Very strong general programming skills (non-machine learning). I’ve spent the last 4 years focused on becoming an excellent developer. If you ever don’t understand something about a programming language, or are not sure how to implement something, send me an email and I can probably get you pointed in the right direction.
Comments
It’s been hard enough to find one other person interested in AGI, so the field has seemed out of reach. I’ve been hoping that a job or conference would help break me into the academic community. Watching the live stream and hearing, essentially first hand, about the bleeding edge theories has made contributions to the field feel accessible and tangible, which is fantastic. I look forward to uncovering the mystery of intelligence with you all.
The reasons that made you come here are precisely the same as mine. There is no AGI out of brain simulation either in software or in hardware, and I think that most of the ML schools of thought are nothing more than very good statistical methods to deal with high dimensional data. Numenta seems very promising as an approach towards AGI (in my opinion). I am also first studying all that I can about HTM before getting into coding. I am doing this it’s been two weeks. From my (yet short) experience I think you’re doing the right thing throwing yourself into HTM School. Have a good journey.
Sadly, I have not yet dived deeply into the HTM School. I’m trying to learn it slowly it on the side.
Took Andrew Ng’s Coursera machine learning course 4 years ago.