ivan.moony

Artificial God? in General AI Discussion

Sentient beings seem all about input/output. A being perceives some input and responds to that by some output. Then it is all together fed back to input, which generates some other output, and so on, while the being lives.

The current situation is that we have a way to train artificial neural networks on supercomputers that can be interpreted on average computers, providing an output for a certain input. But certain outputs are unacceptable, or ethically questionable, so it wouldn't be wise to always pick the first output from the stack. The solution might be in programming a counterpart to ANNs, which we could call "God". The God's purpose would be to judge if a certain output will be performed, or another output would be requested to judge about.

Now we have two entities to worry about: a sentient being and a God. The sentient being seems more-or-less solved, it could be simulated by ANN, but the God... That should be a real complication, shouldn't it? The question I would like to rise is: "How to program an universal God that handles any ANN trained instance?"

frankinstien

Concept Modeling in General Project Discussion

I'm working with concepts as data expressed as a structure with properties and/or methods. I came up with some basic math shown below:



Where concepts are the elements A,B,C,F,G,Q,Z, and T.

As shown in the diagram there are five basic operations along with the use of some set functions of union and intersection.

• Relatable to: can be expressed as grades or relevance.
• Indirectly relatable: indicates processing to relate to an element, which can also be viewed as a goal.
• Apply: is a process of utilizing verbs, adjectives, and adverbs that have expressions of properties and methods as well.
• Processes into: is a literal function that can produce a concept, e.g. computing the area of a table.
• not: is the anthesis of features or method outputs

MikeB

Pattern based NLP in General Project Discussion

This is a project I've been working on for a few years. In 2019 I was testing out the theory on Pandora Bots, and this year I'm converting it to C/C++.

The main goal is to be small, fast, solid-state. Not algorithms, deducing/knowledge, learning...

It works by matching singular words, first through spell check, then into another full list of words which assigns each word a token (one physical character/byte symbol). There are approx 20-30 groups of symbols which all words match into. The 20-30 symbols are then used in pattern sentences. There are a few hundred pattern sentences. From each pattern sentence a broad intention can be gathered and this used in chatbot reponses.

The Chatbot responses are fixed to around 50-100 and are represented as a one character/byte symbol. These are used to create duplicate / cross language responses. They can also be voice recorded and assigned to each symbol.

It's aimed at time/space restricted Chatbots, for example in games, where processing needs to happen in milliseconds, but also take into account broad user input.

Size and Response times:
In 2019 in pandora bots (1000 words, 200 sentences) the size is ~500kb, and ~1 second response time.
In 2020 in C/C++ on an Arm Cortex M4 @ 120mhz it's ~159kb and 15-100 milliseconds.
In 2021 in C/C++ on a modern PC @ 2.6ghz with Binary Searching (set up time of 70ms), processing is ~1ms / sentence.

A few features:
 < 500kb including word databases.
 One sentence generally takes less than 1 ms to process.
 Limited chatbot responses make it easy to voice record and/or change personality.
 Private information stripped during word compression (no names/places).
 Fine differentiation of intentions, eg between: Wondering, Questions, and Directions - "can you speak english" "do you speak english" "speak english".
 Can count occurance of emotional words, logical words, burning-analyser words, light-sense words to reply in kind better.

For the problem of chatbots in games or time/space restricted platforms, it solves:
 Too much data or processing power required.
 Cannot change the personality/no personality.
 Cannot change the language/only one language.
 Chatbot escaping the topic due to bad intention reading.
 Chatbot returning bad views / knowledge calculation (only fixed responses allowed).
 Chatbot terrible voice synthesis (can pre-record all fixed responses including randomised duplicate recordings).
 Not being white box/solid state/predictable.

yotamarker

java/kotlin to python in AI Programming

is there an online service that can do that
or maybe a prewritten class to translate java/kotlin to python code ?

frankinstien

Quantum computed virtual reality or even computed quantum reality? in General Chat

I had a strange dream last night. I dreamt that our reality was some kind of quantum computing Time-crystal formalized in some icy solid cold comet or even some super cold metal! Where time-crystals created a virtual universe from the lattice of atoms of super cold material that interfere with one another, creating computational chaos that formalized into our reality! Not only that but the comet or asteroid was nearing its star and heating up.  The expansion of the super cold material as it warmed is what is causing the universe to inflate! The lattice of atoms is expanding causing localizations of virtual computed matter to move further apart. So, our existence is short, but because our reality is being computed at the speed of light or perhaps even faster than the speed of light, due to entanglement, to us this short-lived life of our universe will be 100's of billions of years!

So, I wake up and you guessed I looked up whether time-crystals could come about in nature. Turn's out they might, but the article explores classical processes that exhibit time-crystal effects. Given that solid cold matter exists in the Kuiper belt could such structures produce some kind of computational reality where such lattices of matter that move very little and the time-crystal phenomena manifest on a scale of astronomical proportions, that converge into a reality?

But then again, could all of reality be based on this non-entropic phenomenon of time crystals. Where the foundation of reality is some form of medium where elements interact chaotically but eventually form into an organized system that doesn't need energy! In fact, the concepts or phenomena of mass and energy are computed from a very highly parallelized system of stuff that's smaller than subatomic particles. The foundation of reality is without energy, icy cold, absolute zero, but these kinds of time crystals can still interact and converge into a universe.

Realize that all quantum matter has that strange behavior about it, superposition, entanglement, parallelized random state interactions, with strange motions that make no sense to mere mortals. So, if all matter inherits those properties and that is the case, then perhaps a simple model of stuff could be generated and that pattern begins to re-enforce itself by the inevitable chance, given an eternity, that such a pattern then spreads to a critical state that creates a universe. We already have proof that well-organized computational systems can arise out of chaos, they're called brains.

frankinstien

Dendrite Processing in General Chat

Here is a very interesting and fairly recent paper on the morphology and signaling of dendrite spines.

The diagram below is from the article:



Looking at the ion pumps on the plasma membrane and the endoplasmic reticulum (ER) membrane, it's clear, that there is an ability to modulate signaling, allowing either amplification or reduction. Not, only that, but the ability to control a neuron's internal states, such as dichotomies that can act in a boolean fashion where the spine's states can cancel out or re-enforce each other within regions. The paper states that the spines can physically change and that has an effect on the plasticity of neurons.

The beauty of this kind of approach is by simply anchoring to these spines you can build combinational logic astronomically.  

Here's an older paper on the effects of caffeine on dendrite spines.

LOCKSUIT

New cryopreservation method is much better in General Chat

What can be cryopreserved better?:

"New computer model is a key step toward low-temperature preservation of 3D tissues, organs"

"Medical science is a key step closer to the cryopreservation of brain slices used in neurological research, pancreatic cells for the treatment of diabetes and even whole organs thanks to a new computer model that predicts how tissue's size will change during the preservation process."

How better?:

"Healthy cell survival following vitrification rose from about 10% with a conventional approach to greater than 80%."

Ok. Thanks God. I don't want to die. I want patternz. Link:

https://today.oregonstate.edu/news/new-computer-model-key-step-toward-low-temperature-preservation-3d-tissues-organs

P.S.: All you may have to do is store your brain's natural fluid in a jar to put back in after you'd thaw; don't throw nothing out. Also your computer, as it stores a lot about you.

LOCKSUIT

GPT-3 is public in General Chat

https://openai.com/blog/api-no-waitlist/

i think codex is not available, and neither is dalle and jukebox even on yet BTW I think.

BUT I TRIED GPT 3 eeeeeeeeee weeeeeeeeeee and i just click the button and it added more text and more and more and more and was FAST OMG and got this thing:


--------------------------------------------------------------------------------------------------


Cryonics still has some problems, ice crystals form which cause damage and the chemicals to stop that also are toxic to cells. What should we do to make cryonics work better for life extension of the human brain? Here is how:

Improve the freezing process

In order to have a successful freezing process, we need to decrease ice crystal size. If we can decrease ice crystal size, we can decrease the amount of damage.

Cryopreservation process

The cryopreservation process can be simplified. We need to have a way of quickly cooling the brain without ice crystal formation. The current method of cryopreservation is very slow, there are many steps, and each step introduces the possibility of ice crystal formation.

The brain is very difficult to cool and thus we need a way to quickly cool it and stabilize the temperature. This is a problem. If we can do this we can have a much better chance of success.

The current process is as follows:

Agitation

The brain is taken out of the body and infused with a fixative. This fixative is a chemical that inhibits ice crystal formation. This is a problem, because this step introduces the potential for ice crystal formation.

The brain is then agitated to try to remove as much of the fixative as possible, while not damaging the brain. This is a problem because agitation introduces the potential for ice crystal formation.

Ice removal

The brain is then frozen with the fixative. This process is slow and introduces the possibility of ice crystal formation. The brain is then frozen with dry ice. This is a very slow process and the brain is constantly being warmed, introducing the possibility of ice crystal formation.

A new process

To solve this problem, we need a way to quickly cool the brain. We need a way to introduce a solid material that will take a very long time to melt. We need to cool the brain so quickly that it does not have time to form ice crystals. We also need a substance that will not cause any damage to the cells.

There are some very good materials that would work well for this. One of the best substances that I have found is glacial acetic acid. This is an acid that is very slow in melting. If we freeze the brain in this substance, the brain would only take about 20 minutes to freeze. This would eliminate ice crystal formation. It would mean that the brain would be damaged very little or not at all. It would mean that the brain would be in a very good condition to be revived.

The process would be as follows:

Microsurgical removal

The brain is microsurgically removed from the body. This process is not perfect and it is possible that some brain cells could be damaged.

Fixation

The brain is infused with the glacial acetic acid. This process is very fast and there is no possibility of ice crystal formation.

Cooling

The brain is cooled to a temperature just above the freezing point. For this process, we need to cool the brain very quickly, in order to prevent ice crystal formation. The brain is then cooled to -20C. This process takes about 20 minutes to complete.

Cryopreservation

The brain is frozen. The cooling process is very fast, so there is very little time for ice crystal formation. The brain is frozen with dry ice. This process is very slow, so there is time to melt any ice crystals that may be formed.

The brain is then frozen in liquid nitrogen.

This process is much better than the current process. It can reduce ice crystal formation and it can reduce the damage to the brain cells.

This process would work better than the current process and it would decrease the damage to the brain.

This process would be much better than the current process and would decrease the damage to the brain.

The future



Cryonics needs to be changed. It needs to be improved. Even if we do not succeed in restoring the brain, we would still be able to extend the lives of people and their families. This would give them more time to spend with the person that they loved. It would be a very sad thing that we could not restore the brain, but we can still extend life for many people.

If we can improve cryonics, we would be able to extend life for many people. This would be very important for the future of humanity.



https://ibb.co/9nwN1qT



2nd result:

Cryonics still has some problems, ice crystals form which cause damage and the chemicals to stop that also are toxic to cells. What should we do to make cryonics work better for life extension of the human brain? Here is how:

1. Alcor's method of perfusing the brain with cryoprotectant chemicals is a good start. It is basically a freeze-drying method. However, it would be better to perfuse the brain with liquid nitrogen at a low temperature before freezing. This would prevent ice crystal formation and stop further damage. This procedure is called vitrification and is used to cure cancer and other diseases and to preserve and transplant organs. (See the article on cancer research in the September-October 2002 issue of this newsletter).

2. The perfusion fluid should be two-phase liquid nitrogen and liquid nitrogen should be used to replace the water in the cells in a gradual manner. This is called vitrification in cryobiology and it is a way of preventing ice crystal formation in tissue. The cells are frozen with a very high concentration of cryoprotectant in the outside of the cell and a high concentration of cryoprotectant in the inside of the cell. This causes the cell to freeze in a glassy state without ice crystal formation. It takes a long time to pump into the cell, but is better than freezing. This is the only way to prevent damage from ice crystal formation.

3. The preservation solution should be a simple solution of a few chemicals, not a cocktail of a dozen different chemicals. This would make the freezing and thawing easier and reduce the chance of toxicity. It would also allow for a whole body transplant into a new body after a few hundred years. The use of a few chemicals that are easily replaced by the body is called a Biostasis procedure and is used to preserve organs for transplant. (See the article on organ transplants in the September-October 2003 issue of this newsletter).

4. The patient should be placed in a large thermos bottle, not a small cryostat. This would give the brain a better chance of not being damaged from ice crystals. In addition, the body should be placed in a thermos bottle filled with liquid nitrogen, not just the head.

5. The cryonics organization should have a backup of the cryoprotectant chemical solution and a backup of the body. This would allow the cryonics organization to continue cryopreserving people even if there was a major earthquake or other catastrophe.

6. The cryonics organization should have a backup of the patient's brain. This would allow the brain to be thawed and transplanted into a new body in the future.

7. The cryonics organization should have a backup of the DNA of the patient. This would allow the DNA to be cloned and the cloned person to be brought back to life in the future.

8. The cryonics organization should establish a set of procedures and a set of rules and regulations for cryonics that are clear and precise. This would avoid problems with insurance companies, government, and lawsuits.

9. The cryonics organization should have a backup of the DNA of the patient's children. This would allow for the DNA to be cloned and the cloned children to be brought back to life in the future.

10. The cryonics organization should have a backup of the DNA of the patient's grandparents. This would allow for the DNA to be cloned and the cloned grandparents to be brought back to life in the future.

11. The cryonics organization should have a backup of the DNA of the patient's great-grandparents. This would allow for the DNA to be cloned and the cloned great-grandparents to be brought back to life in the future.

12. If the cryonics organization is in the United States it should have a backup of the DNA of the patient's ancestors back to the days of the Pilgrims. This would allow for the DNA to be cloned and the cloned ancestors to be brought back to life in the future.

13. The cryonics organization should have a backup of the DNA of the patient's relatives. This would allow for the DNA to be cloned and the cloned relatives to be brought back to life in the future.

14. The cryonics organization should have a backup of the DNA of the patient's pet. This would allow for the DNA to be cloned and the cloned pet to be brought back to life in the future.

15. The cryonics organization should have a backup of the DNA of the patient's brain. This would allow for the DNA to be cloned and the cloned brain to be brought back to life in the future.

16. The cryonics organization should have a backup of the DNA of the patient's memories.

frankinstien

Natural Language Cognitive Architecture in AI Programming

LockSuit pointed me to this book "Natural Language Cognitive Architecture" by David Shapiro:



The book focuses on natural language generative Transformers that Shapiro hails as a milestone in AI. He extrapolates transformers to a generalized AGI described in concept by the diagram shown below:



Where the outer loop is a context-driven process that has inputs from external influences of an environment. And the inner loop is a form of streaming consciousness that can reflect on influences of the outer-loop as well as the inner loop's inferences and both share a common database. The book doesn't go into too much detail in terms of architecture and he states that current approaches with transformers are best suited for single-threaded approaches. He also uses SQL relational databases. While the generative transformers are able to confabulate and/or extrapolate contextual patterns from prompts and cues that can provide a context or goal, they still suffer from codifications that are cumbersome because of how neural networks code information. The whole approach starts at such a small datapoint knernal which are the characters of a language! Then the ANNs build up structural elements within its layers that allow for identifying concepts. The idea of ANNs being burdened with such data, IMO, is an inefficient approach to using data and ANNs!

An alternative approach would be to remove the burden of actual structural data of concepts and use databases to provide that resource. But a SQL database is the wrong approach. Why? Relational databases were designed to remove ambiguity and as such form relationships of data that are very brittle. Tables have fixed columns and language deals with data types that will vary in context data and fields. That's a problem, to make relational databases more flexible you'd have to use join tables to represent varying fieldsets for different concepts. So you'd have a concept table that points to a field table. The problem with that is you now have to do very computationally expensive joins that become a nightmare to associate across varying ideas and contexts.

A better approach is to use an object-oriented data model. Looking at the diagram below:



We can see a word or concept is described with vectors that need to be exposed so associations and comparisons can be made. This approach thrives on parallelism because we can build hash sets of the feature vectors that provide a time-complexity of O1 lookup advantage where 100s of thousands of lookups can happen in parallel! This removes the need for an ANN to form codifications that structure data into relationships. Those feature vectors include grammar, context, and a plethora of other concepts which would allow for an ANN to do a much simpler process and that is use the structured data and functions that can search, compare and process by virtue of patterns of using external functions rather than having to perform them. So, take for example GPT-3 learned to add, but why learn to add if it's a functional process a machine already can do and is coded much more efficiently than an ANN to do? Wouldn't it be better to train an ANN to use a calculator rather than form internal logic to do arithmetic?

Now look at Amanda's AGI approach:



The diagram depicts a concept of time that has varying degrees of depth, this is very similar to what human brains do. Where Shapiro obsesses with time-stamping data that effectively turns events into points of time which is not as effective as what nature invented. Time depth gives us a sense of work-effort which biology relies on to conserve energy. Not only that but the architecture of Amanda implements parallelism and cross-platform capability so when it needs to use a GPU, it can and that function has descriptors that describe it like any other concept. This allows for the search for a function to be easily associated with the concepts it processes by virtue of hash set lookups of feature vectors!

Training an ANN to use functional capabilities, such as looking up data, comparing for appropriate word use and document evaluation removes the need for knowledge as stored information to be solely an ANN responsibility and focuses the ANN to find patterns of efficacy using functional processes and externally(external to the ANN) stored data! With this approach, there's no need to re-train or fine-tune for the memorization of new data. Nor does this approach require an ANN to learn gigabytes of data, it learns spontaneously by interacting in its environment and can grow its vocabulary and experiences but relies on other processes to do the heavy lifting of comparing, weeding out, and assessing best fit data.

At least that's the theory, let's hope I'm right...

frankinstien

My HAl Rig in General Hardware Talk

Since I believe the neural network approach isn't the most efficient use of digital resources because it effectively is a linear search in a problem or knowledge domain, no matter how much pruning of a network you do. So I'm using a good deal of RAM, 128GB, and two 10 core Xeons. This doesn't mean I don't use GPUs, I do and its a Vega 56, I use it primarily to work with fuzzy logic and fast index calculations as well as process video and audio data. Below is a diagram of how I organize memory from RAM, SSD to hard drives.



The SSD is a 1 TB drive and it stores data from a noSql database I engineered that uses O(1) lookup indexing. The Hybrid hard drive is more of a backup of the SSD but will store very large datasets like long video. You'll notice a block called "Cross Domain Harness" I really wanted to avoid a lot of serializing and deserializing between processes, so the harness allows the dynamic injection of various programs under a single domain. Since when I prototype I haven't necessarily integrated all the pieces yet and I don't want to deal with the latency of serialization. This way I can pass data by reference so long as the programs have imports to the datatypes used. It makes life a lot easier to troubleshoot particularly when the amount of data loading is in the 10s of GBs.

The machine was built three years ago so upgrading is starting to seem like a good idea since AMD's processors are reasonably priced, however, the RAM prices haven't dropped by much. So if I did upgrade it's a 50% improvement in performance using the Ryzen 9 3900X 12-core. Because my current machine is a dual-socket Xeon board it has 16 memory slots which allows me to use lower-cost 8GB sticks. In fact, I do debate if it's worth putting up the money for the newer Ryzen or just getting 16GB sticks to work with larger datasets?