Friday, March 15, 2024

AI R Us

This should not be surprising.* AI models are a mirror to human behavior. So shit that goes in results in exporting shit. 

AI models are trained to learn, consolidate, collate, and regurgitate. We don't like what we get, so we change it afterwards. But that doesn't change the mirror effect. It's a Band-Aid approach: Cover the abscess so we don't have to see it. Or, as Nikhil Garg, a computer scientist, eloquently puts it, “simply paper over the rot”.

This brings to mind a sci-fi series I'm watching on TV, "Beacon 23". AI, a common component of most sci-fi, is represented as two important components in the series, and are almost complete opposites: "Bart" and "Harmony". The latter is a logical Spock attached to an individual human and created by a major company that controls almost everything.

Bart is an AI developed hundreds of years prior to the primary time event of the series and serves the structure, Beacon 23, in space (it's a 'lighthouse' stationed near a flux of dark matter.  The Beacons guide space craft away from detected nearby dark matter 'clusters', like a lighthouse.) 

Bart learns and adopts human nature/behavior from all the beacon "keepers" and those who visit. He also controls all the communication and mechanics of the Beacon station. As such, Bart is just like a human, with all the messiness, assumptions, errors, etc: he lies, whines, plots, complains, quotes Shakespeare, and often acts like a child.  But Bart also significantly sets the course for what occurs on the Beacon. Whereas Harmony is logical and attuned only to her individual human, but also capable of Beacon control, including Bart (she scolds Bard many times).

Humans created AI, and in its current state in our world is like a young Bart. That we can't see that is human blindness. AI is not, and won't be our savior. We can't even save ourselves. 

"Even though human feedback seems to be able to effectively steer the model away from overt stereotypes, the fact that the base model was trained on Internet data that includes highly racist text means that models will continue to exhibit such patterns."

* "Chatbot AI makes racist judgements on the basis of dialect," Elizabeth Gibney. Nature, March 13, 2024.

Monday, March 11, 2024

Which Reality is Real?

What is 'reality'? Ask 10 people that question and you'll likely hear 10 different answers, including "I don't know." The question may be as old as human consciousness. However, it may not be a realized 'thing' [1] that all humans ponder. 

A quick summary in the online encyclopedia Wikipedia present reality as..

"Reality is the sum or aggregate of all that is real or existent within the universe, as opposed to that which is only imaginary, nonexistent or nonactual. The term is also used to refer to the ontological status of things, indicating their existence. In physical terms, reality is the totality of a system, known and unknown."

Is reality a thing? Or is a way of looking at 'things'? 

Reality may be both. No one may have understood this more than physicist Erwin Schrödinger, famously known for his thought experiment (1935) in quantum mechanics, Schrödinger's cat, . It was as an argumentum ad absurdum (reductive argument to absurdity) intended for questioning the then proposed behavior of atoms and larger manifestations as being one or the other, as in "dead or alive", and which depends on the observer. Or, simply put, it suggests that reality is relative to the organism that observes or experiences it in one way or another.

Recalling an old platitude: If a tree falls in the forest and no one is looking or hearing it, did it really fall?

 Yet, Schrödinger's paradox legitimately questioned, 

"When does a quantum system stop existing as a superposition of states and become one or the other?" (More technically, when does the actual quantum state stop being a non-trivial linear combination of states, each of which resembles different classical states, and instead begin to have a unique classical description?"[2]

More simply put, 

"Our intuition says that no observer can be in more than one state simultaneously—yet the cat, it seems from the thought experiment, can be in such a condition. Is the cat required to be an observer, or does its existence in a single well-defined classical state require another external observer?" [2]

This may bring to mind Einstein's "Theory of Relativity," which is associated with quantum mechanics. Indeed, Einstein considered each alternative as absurd. He wrote to Schrödinger,

"You are the only contemporary physicist, besides Laue, who sees that one cannot get around the assumption of reality, if only one is honest. Most of them simply do not see what sort of risky game they are playing with reality—reality as something independent of what is experimentally established. Their interpretation is, however, refuted most elegantly by your system of radioactive atom + amplifier + charge of gun powder + cat in a box, in which the psi-function of the system contains both the cat alive and blown to bits. Nobody really doubts that the presence or absence of the cat is something independent of the act of observation."[3]

Many interpretations, both technical and popularized,  provide explanations and answers to Schrödinger's paradox. However, the quantum world is full of counterintuitive ideas, which was strongly implied in Schrödinger's thought experiment. Several physicists contemporary with Schrödinger and after his passing proposed their own perspectives. 

Of note, American physicist, Hugh Everett who  proposed (in his 1957 PhD thesis) what is now known as the many-worlds interpretation of quantum mechanics. I'm sure most readers here are familiar with the popular "multi-verse' of science fiction genre and even modern physics. Everett's idea of quantum mechanics does not single out observation as a special process. His many-worlds interpretation of Schrödinger's paradox explains that, 

"...both alive and dead states of the cat persist after the box is opened, but are decoherent[4] from each other. In other words, when the box is opened, the observer and the possibly-dead cat split into an observer looking at a box with a dead cat and an observer looking at a box with a live cat. But since the dead and alive states are decoherent, there is no effective communication or interaction between them. 
When opening the box, the observer becomes entangled with the cat, so "observer states" corresponding to the cat's being alive and dead are formed; each observer state is entangled, or linked, with the cat so that the observation of the cat's state and the cat's state correspond with each other. Quantum decoherence ensures that the different outcomes have no interaction with each other. The same mechanism of quantum decoherence is also important for the interpretation in terms of consistent histories. Only the "dead cat" or the "live cat" can be a part of a consistent history in this interpretation. Decoherence is generally considered to prevent simultaneous observation of multiple states."[5]

Quantum mechanics is often used in contextual explanations of reality. One could possibly, and loosely, refer to Everett's hypothesis as 'alternate' realities. Is this a real 'thing'? Or just another thought experiment or interpretation of reality?

This all begs the question, is there just one reality? If so, then what is it? Afterall, a 'real' reality could exist for non-living things (we know they exist, but the non-living have no consciousness), and another for living things (because we have consciousness and are aware). Which suggests that an infinite number of personal realities may exist. A shared reality may be then be overlapping personal realities like many flexing Venn 4-dimensional boxes that overlap, constantly shifting, temporally and spatially. 

As mentioned earlier, reality is the the totality of a system with known and unknown existences.

Or perhaps reality is like the smile of the Cheshire cat: it remains even when the cat becomes invisible.
_______________________________________________

[1] I'm compelled to explain my use of 'thing' as in it's most recent etymological context: used colloquially since 1600 AD, as a word to substitute for what a person cannot think of it's meaningful name. However, this ubiquitous word 'thing' has an interesting history back to the Vikings. 
[2] "Schrödinger's cat," Wikipedia
[3] Letter to Schrödinger in 1950.
[4] In quantum physics, decoherence is the process in which a system's behavior changes from that which can be explained by quantum mechanics to that which can be explained by classical mechanics.
[5] Hugh Everette III, Wikipedia. As an aside, both Hugh and his son, Mark Oliver Everett, are thought by many to be Asperger's (on the autism spectrum).