All must be open-ended. The learning agent (the ) must have the to set its own learning goals as well as plan and execute a of activities to achieve these goals.

One can never learn all existing data but rather refine their understanding of the data that is available to them. As true for human intelligence, you can either have “deep and narrow” specialized agents or “average and broad” . You can’t have both in the same entity. Time and “limitation” are the main inspirations for and between learning agents.

People should have figured it out by now that the of processing power, not the in gargantuan data and control centers is the right thing to do.

Stop working on LLMs (Large Language Models) and start working on PCAs (Personal Customizable Assistants).


The current craze over social media ruining democracy and AI posing an existential threat to humanity stems from the fact that most people don’t understand that the only thing is able to do is amplify their own capacity to do good or bad.
The Internet and AI are communication and intelligence , the same way motors and servo mechanisms are amplifiers of our muscle power.

, like is not a thing. It is a or property of the part of an entity’s (which is a thing) that the entity may be conscious or aware of … or not.

Furthermore, consciousness is a proposition. One can be either conscious of something or not. You can’t be a little bit conscious the same way you can’t be a little bit alive.

I read a sample of Robert M. Sapolsky’s new book Determined: A Science of Life without Free Will on Amazon, and I really don’t see why some people find it “revolutionary”. I find it full of half-baked contradictory claims that don’t hold water even under quick superficial scrutiny like this.

Brains don’t generate behaviors. They motor responses to sensory stimuli that an outside observer then interprets as the behavior of the observed individual in their immediate environment. The observer can also stick electrodes in the brain of the individual and then correlate the observed behavior with the measurements performed on some of the neurons and then conclude that those firings have caused the behavior. However, even if it was possible to replicate the exact sequence of the observed firings of all the neurons, the observed behavior would be different if the “response” of the environment was also not exactly the same as during the measurement.

Determinism alone doesn’t “cause” anything even if there are no such things as “causeless causes”. The current of the system is obviously determined by its previous state and the current sensory inputs, so there are at least two separate “determinisms” in play here all the time, and, as an individual existing in its particular environment, I have at least some over the unfolding of both, my biology (eating, drinking coffee), and my environment (writing this nonsense)😉.

I wish people who are coming up each day with a new “breakthrough” theory using physics and/or quantum mechanics to explain everything from complexity and life to consciousness and free will, would read first what has said about it.
This is from:–the-Modern-Mind-1965-ocr.pdf

This is so wrong I wouldn’t know where to start.

Hurricanes are ‘selected’ based on their ability to perform functions dictated by the environment, the researchers found.

Gimme a break!🤨

in “Life’s Irreducible Structure” (1968) points out that using deterministic to explain “the physics of ” is backward thinking, because machines are devised and built by humans to resemble organisms and to serve the purpose of their design, and can therefore only be a , not a analogy.

The organism is shown to be, like a machine, a system which works according to two different principles: its structure serves as a boundary condition harnessing the physical-chemical processes by which its organs perform their functions. Thus, this system may be called a system under dual control (emphasis mine). Morphogenesis, the process by which the of living beings develops, can then be likened to the shaping of a machine which will act as a boundary for the laws of inanimate nature.

In the machine, our principal interest lay in the effects of the boundary conditions, while in an experimental setting, we are interested in the natural processes controlled by the boundaries.”

Or in other words, we are interested either in the control of the machine or the physical of causality that make the machine work.

Show thread

Wiener was wrong. There is no in either the animal or in the machine, only by the application of to their inner flow of matter and energy.

Communication is established between animals and/or machines, and, as Shannon correctly recognized, requires an independent communication susceptible to the environmental disturbance called .

In order to be able to communicate animals and machines must share a common or cipher used to code their respective messages. Communication is always one-way and does not require feedback. The sender has no control over the message after it is sent through the channel.

A special case of communication is where the communication is established between the observer system and phenomena in its environment not necessarily produced by other systems “languaging”.

A dynamical system with with the ability to learn and adapt to its environment or to change it will need at most these three mechanisms:

1️⃣ The immediate control () of state variables essential for preserving the stability or of the system. This is a simple of the system to a perturbance, like, for example, sweating when the core temperature of the body increases beyond some preset margin.

2️⃣ The control of the surrounding environment is used when 1️⃣ is overwhelmed and there is a need for the coordinated engagement of different lower-level regulators, the (tracking), and negative control of multiple time-dependent variables like for example, when taking off layers of clothes, moving the body into a shade, or taking a cold shower until the temperature gets again within limits.

3️⃣ The , long-term, open-loop control with delayed feedback is the highest form of control, like for example when building a house with an HVAC system that will remove the necessity for a continuous employment of proximal control (2️⃣) by creating a private controlled environment.

All systems feature this 3-layered control architecture, with the only difference being in what degree the activities on each level are the result of deliberation as opposed to a natural, innate behavior.

The thought experiment in this interesting 2019 article from Michael Lachmann and Sara Walker on the contrast between and is not representative, because von Neumann’s UCs are non- and don’t print themselves so can’t and .

Imagine you have built a sophisticated 3D printer called Alice, the first to be able to print itself. As with von Neumann’s constructor, you supply it with information specifying its own plan, and a mechanism for copying that information: Alice is now a complete von Neumann constructor. Have you created new life on Earth?

The difference lies in the fact that UC “mechanisms” are not operational until their production is fully finished and any will most probably prevent the mechanism from working, while most living and growing “assemblies” can work and repair themselves while they are growing.

The bottom line is that life cannot be created. It has to emerge from mechanical non-life.

People often “blame” Shannon’s theory of for completely ignoring , maybe also because Shannon himself stated that “the semantic aspects of communication are irrelevant to the engineering aspects“😀

However, if one recognizes that the content as defined by the is the measure of in a receiver about the sender’s when producing the message, can it perhaps be interpreted that the receiver is trying to what the sender was to send?

The information the sender encodes in the message is never the same as that the receiver decodes from it on the other side of the channel.

Below is Shannon’s description of the standard used for encoding and decoding the information in messages. The block diagrams are my rendering of the description (F is a “” function):

as “deliberate and ” should be easy to explain, but only by the conscious agent itself. There is no way an outside can identify if an agent’s behavior is conscious or not.

— what seems “ and ” — may not be . It may seem “obvious and undeniable” to someone interacting with ChatGPT that it is communicating with a conscious agent, but that assumption would be flawed.

Anil Seth thinks “Conscious AI Is a Bad, Bad Idea” because

our minds haven’t evolved to deal with machines we believe have .

On the contrary, I think we are “genetically programmed” to ascribe intent to anything that “wants” to communicate with us.

He is also saying that:

Being intelligent—as humans think we are—may give us new ways of being conscious, and some forms of human and animal may require consciousness, but basic conscious experiences such as pleasure and pain might not require much species-level intelligence at all.

If, as he says, “intelligence is the capacity to do the right thing at the right time,” any organism that has survived long enough to procreate must have some kind of intelligence, regardless of its consciousness.

Wrt “basic conscious experiences such as pleasure and pain,” IMO they are conscious only if the organism is intelligent enough to suppress the urge of an innate “genetically programmed” response to pain or pleasure in order to achieve some “higher goal,” even if it goes against the original goal of “to survive.”

The bottom line is that consciousness is not just a function of intelligence. Machines can become much smarter than us without becoming conscious.

In order to be really , a machine would first have the experience of being and the desire to remain in that state, have some and over its internal and external states, the ability to develop short and long-term goals and plan and execute complex time-dependent actions to fulfill those goals.

Anything less than that is just a clever simulation.

The purpose of is to new or/and different structures (artifacts), so speaking of design makes sense only if it is in the context of other creative activities such as writing, painting, engineering, manufacturing, etc.

Klaus Krippendorff has a nice description of the difference between and and the relationship to Gibson’s in this 2007 paper published in “Kybernetes”:

However, he is wrong, IMO, in accentuating the difference between and .

Every designer is often a scientist in “describing what can be observed” and every scientist also has to design new hypotheses, theories, and experiments for the “not yet observable and measurable”.

A few “gems” from on the “accumulation of adaptations”:

“A compound event that is impossible if the components have to occur simultaneously may be readily achievable if they can occur in sequence or independently…
Thus, for the accumulation of adaptations to be possible, the system must not be fully joined.

The idea so often implicit in physiological writings, that all will be well if only sufficient cross-connexions are available, is, in this context, quite wrong.”

I recommend reading the whole book ($20 on Amazon) but if not, here is a good overview of some interesting parts:

Show thread

in his “Design for a Brain” writes about the importance of the of . Following his ideas I’ve made this little experiment using a LibreOffice Calc spreadsheet that shows three different scenarios:

When re-tossing all of the 10 coins every time like in the first case there is no preservation of “1s” whatsoever. Every new toss starts from scratch.

In the second case, each coin is tossed separately until it shows “1” when the tosser moves on tossing the next coin until all 10 show “1” which usually happens around the 10th tossing.

In case #3 only the remaining “0s” of the previous toss are re-thrown until all coins show a “1” which is by far the most efficient way of preservation, needing less than half of the time and ending in about 4 tosses.

Our brains evolved as control mechanisms for the body to ensure its survival. That’s the reason why is beating humans primarily in areas requiring computation and abstract reasoning. We only recently added those to our tool repertoire and didn’t have too much time to perfect them as for our ancient sensory-motor control tools.

“Encoded in the large, highly evolved sensory and motor portions of the human brain is a billion years of experience about the nature of the world and how to survive in it. The deliberate process we call reasoning is, I believe, the thinnest veneer of human thought, effective only because it is supported by this much older and much more powerful, though usually unconscious, sensorimotor knowledge. We are all prodigious Olympians in perceptual and motor areas, so good that we make the difficult look easy. Abstract thought, though, is a new trick, perhaps less than 100 thousand years old. We have not yet mastered it. It is not all that intrinsically difficult; it just seems so when we do it.”

Moravec, H. Mind Children: The Future of Robot and Human Intelligence. (Harvard University Press, 1988).

as reported in:

There is a lot of talking against “ thinking” and planning in and how the only “good” hierarchy is a “flat” one.
The above diagram shows how a is a natural effect of folding the linear “information ” to match the physical structure of the system.

For nearly 4 decades in organizational change management, I’ve been using this idea of layers emerging from the of a sequential “information processing” and just found that a whole area in biology deals with this exciting topic.👇

Show thread

“An algorithm solves a problem only if it produces the correct output for every possible input — if it fails even once, it’s not a general-purpose algorithm for that problem.”

Most commenters do not realize that no “information processing” ( on symbol sequences) of any kind is necessary for an agent to have over their internal and surroundings.
Think of a or comparator such as in (Perceptual Control Theory).

Show thread
Show more
Qoto Mastodon

QOTO: Question Others to Teach Ourselves
An inclusive, Academic Freedom, instance
All cultures welcome.
Hate speech and harassment strictly forbidden.