Is Avaama an AI Company? What about Lumiata? Is it enough just to add .AI to your domain name to get the big bucks?

Companies like Avaama are a bit hard to understand. with 28 million in funding from Series A from the likes of Intel Capital and Wipro, really all they are doing is putting a speech query system in front of standard applications. Is this AI? Not really.

There is a lot of talk about making speech better and many systems have advanced features. MaryTTS is one good example where you can even specify moods like Happy in the response. Is this AI? Nothing is being inferenced.

Noonean's first product - NeuralBase - thinks about what data you are asking for and uses several deep analytics which are AI based including neural network language systems to provide more sophistication rather than simply speech to text.

Noonean's second product - OnticBase - extends that with knowledge management capture into a neural structure. This provides even more information.

But we still have a lot of work to do. It used to be that seed round was initiated with a clear product concept and business plan. Now many people want to see beta level products. This was OK with simple websites, it's not so OK when complex AI products are part of it.

Lumiata is another "AI" company getting a lot of investment. They take big data from healthcare and then use "AI" to analyze it. This special sauce is never specified there is never anything described as to their technique and what makes it "AI"

Sadly investors are not savvy and many are going to be fooled with systems which barely could be said to be doing AI but then again, definitionally, what qualified for AI or machine learning. Some people seem to have a very shallow concept of this. They are willing to throw buckets of money on anything remotely having anything to do with AI, well hopefully they will be interested in a real machine learning company with real products not just a big data cruncher or a speech to app player.

This is going to be a big problem going forward as more companies jump in with somewhat "MEH" offerings.

What is an Ontic Processor?

A lot of people ask me what is an Ontic Processor, a term we use at Noonean Cybernetics to describe one of our supporting components. I've even had people get huffy and puffy "That Doesn't Exist!". Well yes it does we use it every day.

No it's not a Pregel graph matrix processor or Deep Graph Transition matrix. Those typically require multiple pass point to point traversal.

Instead the Ontic Processor is a bit like a neural network in reverse. The Ontic processor maps terms and associations like language item -- Role -- Connection Role -- Other Item

The way it works is a bit closer to what happens to a human brain when you stimulate it with an electrode. You stimulate several nodes, and then look at what result nodes get excited. It is a complicated dance through BILLIONS of Neurons and connections that is a bit like chaos theory, utterly unpredictable.

Getting the data mapped into an ontic structure is a bit challenging. It requires a bit of step by step learning, and asking if associations and understandings are correct. But it grows ever more savvy over time.

So perhaps we might say it is not so much a database gobbler, a big data cruncher, or a document parser. Rather, it builds interactively, learning and correcting. Once some sufficient linguistic structure (although it encodes knowledge relationships so it is so much more than NLP) then it can begin digesting more information more quickly, and again continuous testing of relationships through questioning.

When we develop cybernetic systems we have typically large processors to handle senses like vision and audition, and another very large scale neural network configured as a Neural Cube  or ANCPHSELAC  ("ancephelac") (Advanced Neo-Cognitron with Pribram Holographic structures and Edelman Live Association and Competition).  This includes a consciousness processing function as well which is gathering and weighting and responding to inputs while "thinking" about them. So to better understand what is going on, the information is passed to the Ontic Processor, and then much more specificity and more possibilities about the language and knowledge is returned, specifically relationships, part composition and decomposition, roles in relationships, and other terms which are not originally in the language. For example, if someone says "Had a problem with the car's rotation" it would get information back that the tires are rotated, rotate, and so does the engine. So it might reply "You rotated your tires or do you mean a problem with the engine" This is much more interesting linguistic interaction rather than the "DO YOU WANT TO DANCE. I CAN DANCE" nonsense. One can only wonder what a Boston Dynamics Language processor will sound like "MOVE AND I KILL KILL YOU". Scary.

So the Ontic processor is a necessary advanced way to retain knowledge but NOT memory. But it can be a supporting classifier for memory and item IDs in memory. That work still to come.

One Terabyte in a postage stamp vs. 16k in a brick ....

You will laugh. This is an atari 800 and this is the very expensive 16k ram expansion which is the size of a brick. Of course this only stores information as long as there is electricity, turn it off and it all goes away. Silly huh? 

And then today we have microSD cards which hold one terabyte, and the data never goes away. which is  62,500,000 times the storage of the brick in case you wondered. Wasn't it AC Clark who said any suffienciently advanced technoloy will appear as       Magic.  To an old geezer like me its better than Hairy Putter.

Integrating Neural Networks with LL(*) Parsers and Natural Language Processing

Natural language processing has been a traditional area of Artificial Intelligence. Transition Networks and Augmented transition networks have defined language grammars since 1970.

Java has taken grammars further with support for Java Speech Grammar Format (JSGF). The advantage of having JSGF is it makes it simple to parse back in results for more analysis.

Like AWK/SED parsers in Unix, JSGF takes the problem of stating a complex grammar further by allowing a straightforward syntax in text to represent a grammar. This opens the possibility of defining a self changing grammer.

The premier parser today is ANTLR4. It has versions for C and Java. But another possibility is the Apache OpenNLP which is much more primitive but provides access to source and some minimal trainability but is mostly a part of speech type old style NLP.

Then there are LL(*) grammars which are more useful in processing continuous speech. There are other techniques to help decide initial branch choices such as Probabilistic Recursive Transition Networks (the A now changed to R to demonstrate the more accurate issue of recursive definition)

But one issue is how to produce a grammer which is more like how Humans learn language. Could you define a infrastructure which itself is self learning and changing to pick up any language, whether German , Swahili, or abstract. That is quite a bit far off in the future.

But one area where neural networks can augment these systems is by providing learning and reinforcement training into weights and modifications of grammars and parsing trees.

One particular area which should have process are the ability to determine response especially to an interrogative. But also in the realm of general speech.

One problem with systems like SIRI and ALEXA is they are obviously rigid and stupid. Much of these systems have been done simply with collections of questions and statistics. Try to ask a reasoning question and they barf. Try to even ask them to convert your weather results to Farenheight and they barf. Why? Because they are not truly modeling language.

There are two primary issues involved. The first is it will require true semantic network modeling in order to provide more intelligent responses rather than a simple grammer with no knowledge representation. This is however a large effort which is going to require a revolution from the image based Neural Network paradigms like Tensor. This is why I invented the Neural Cube which is the next generation of Fukushima's Neo-Cognitron adding concepts from Karl Pribram and Gerald Edelman and John Koza.

Integrative systems will provide processing which can effect path choice weighting which can improve the efficiency of parsing grammars for real time speech or add deeper classifications to grammers - e.g. temporal location vs. noun.  They will also be able to tie in parallel neural networks to inject things like emotion, mode, task participation and state, and also tailoring response based on the relationship with the person. We don't want a cold robotic response like SIRI we want an empathetic response which shows understanding of the situation. This will become the principal feature when developing cybertrons for common human tasks such as working a hotel reception desk. People can be forgiving if the language or the process isn't perfect, as long as they form the empathetic connection. People become frustrated quickly if the effect is dealing with a cold machine, which as I hurl my Alexa device into the wall becomes ever so apparent.

Cybernetics and Cognitive Science stresses to build in adaptability and learning in many ways to make the systems work closer to how our brains work. While we have so much to learn in processing language, this is still an area where basic techniques can greatly improve the state of the art.

Noonean calls this "More natural language". Issues like emotive, goal, variability, and language that humans find pleasant is often more important than being able to answer complex questions correctly. That complex reasoning will come. But, the stoic rigid language of Siri and Alexa need to fade into history as a silly first attempt. It is still revolutionary, almost magic, but what is coming next will be language that doesn't annoy. Language that makes Humans comfortable. This is the promise of cognitive linguistics and cybernetic language development.

Quoted in a Thesis

I feel a bit like I've arrived when other researchers quote me in their work !

Bitter Electromagnet As Foundation Principle for Synthetic Muscles

The nylon twist coil heat activated synthetic muscle at first seemed promising. But ultimatly we could not get a reliable solution. They tended to break and fatigue, adding heat without melting the nylon was difficult. Braided carbon nanotubles was another solution but decades away in mass production as well as being too small.

So our synthetic muscle development is working on a principle which is used in Bitter electromagnets,which is a distributed helical coil field. This provides much of what we wish, a strong stable and efficient response to low voltage pulses.


The key aspect is how to manage contraction and retraction such that these can be durable for millions of pulses. But the unique Bitter design which is used in the worlds most powerful magnets, works just as well on the micro scale. As above so below ....

Using Tertiary Structures in HBM2 Arrays To Accelerate Neuronal Access

HBM2 architecture still suffers from bus latencies across the interposer substrate. Wider memory buses and stacked memory architecture still does not ameliorate the bus differential distance between the GPU processing core (stream processor) and HBM retrievals. This works find for mass bulk data screen loads supporting graphics but is a terrible architecture for neural processing.

Instead an interposed fine grained tertiary structure between HBM and each stream processor can improve performance by a factor of several thousand times for retrieval and update operations.

Nvidia's GP100 uses the same slow approach relying on high wattage bus speeds



The interposer tech is designed for one processor to access large scales of memory whereas the neuronal unit requires localized memory with hyperspeed interconnects which transcend each planar layer. A few random spiking units defer to interposing bus transactions alternating bit transfers and based on non-linear excitation strategies.

Here is a brief overview of correlet architecture for synaptic processing without large bus latency



This produces an inner group latency which is nearly instantaneous while larger outer group branches simply degrade exponentially according to distance. The main complexity with neural cube architecture is transposing non-linear synthetic distances into core relationships as its nearly impossible to dis-assimilate plane relationships which are inherently larger correlet hops than planar distance would indicate.



To ameliorate this situation, a plane dragging event is created to provide sequential ordered plane processing where required. Luckily in true convolution the cyclical stimulus is often non-ordered.

current implementations fail at this accelerated locality. Larger scale designs which operate on stimulation impedus rather than constant high rate bus cycling is the answer, something IBM has achieved with very small scale architecture (256 neural units) while true neural cubes begin to take form at 64 billion neuronal units. To get there synthetic networks transposing to high powered local groups is the optimal architecture. Eventual non-synthetic processing will be ideal, most likely when organic circuits develop.