Distributed Contexts in the Language Game

The meaning of words and phrases can be a bit hard to pin down. Indeed, the meaning of meaning itself is problematical. I can point to a dictionary and say, well, there is where we keep the meanings of things, but that is just a record of the way in which we use the language. I’m personally fond of a kind of philosophical perspective on this matter of meaning that relies on a form of holism. That is, words and their meanings are defined by our usages of them, our historical interactions with them in different contexts, and subtle distinctive cues that illuminate how words differ and compare. Often, but not always, the words are tied to things in the world, as well, and therefore have a fastness that resists distortions and distinctions.

This is, of course, a critical area of inquiry when trying to create intelligent machines that deal with language. How do we imbue the system with meaning, represent it within the machine, and apply it to novel problems that show intelligent behavior? In approaching the problem, we must therefore be achieving some semblance of intelligence in a fairly rigorous way since we are simulating it with logical steps.

The history of philosophical and linguistic interest in these topics is fascinating, ranging from Wittgenstein’s notion of a language game that builds up rules of use to Firth’s expansion to formalization of collocation of words as critical to meaning. In artificial intelligence, this concept of collocation has been expanded further to include interchangeability of contexts. Thus, boat and ship occur in more similar contexts than boat and bank.

A general approach to acquiring these contexts is based on the idea of dimensionality reduction in various forms.… Read the rest

Intelligent Borrowing

There has been a continuous bleed of biological, philosophical, linguistic, and psychological concepts into computer science since the 1950s. Artificial neural networks were inspired by real ones. Simulated evolution was designed around metaphorical patterns of natural evolution. Philosophical, linguistic, and psychological ideas transferred as knowledge representation and grammars, both natural and formal.

Since computer science is a uniquely synthetic kind of science and not quite a natural one, borrowing and applying metaphors seems to be part of the normal mode of advancement in this field. There is a purely mathematical component to the field in the fundamental questions around classes of algorithms and what is computable, but there are also highly synthetic issues that arise from architectures that are contingent on physical realizations. Finally, the application to simulating intelligent behavior relies largely on three separate modes of operation:

  1. Hypothesize about how intelligent beings perform such tasks
  2. Import metaphors based on those hypotheses
  3. Given initial success, use considerations of statistical features and their mappings to improve on the imported metaphors (and, rarely, improve with additional biological insights)

So, for instance, we import a simplified model of neural networks as connected sets of weights representing some kind of variable activation or inhibition potentials combined with sudden synaptic firing. Abstractly we already have an interesting kind of transfer function that takes a set of input variables and has a nonlinear mapping to the output variables. It’s interesting because being nonlinear means it can potentially compute very difficult relationships between the input and output.

But we see limitations, immediately, and these are observed in the history of the field. For instance, if you just have a single layer of these simulated neurons, the system isn’t fundamentally complex enough to compute any complex functions, so we add a few layers and then more and more.… Read the rest

The Twin Earth Dissonance Conspiracy

I came of age with some of the mid-to-late 20th century literature that took conspiracies as truss work for calculated paranoia, from Pynchon’s Gravity’s Rainbow to Philip K. Dick’s identity shuffling, and on to the obscurely psychedelic Illuminati books by Robert Shea and Robert Anton Wilson. They were undoubtedly influenced by the dirty tricks and mind control fantasies and realities of the Cold War, from thallium and LSD poisoning plots against Fidel Castro to the Manchurian Candidate and John Birchers; from Dr. Strangelove to ratfucking in the Nixon-era Republican Party.

The fiction paralleled and mimicked those realities but it was also infused with a kind of magical realism where the ideas permeated through the characters in a nexus of paranoia and fantasy. The reader was admitted to eccentric ways of structuring the history of the world and the motives of unseen forces acting through organizations, governments, and powerful people.

While endlessly fun, the fictional forms were also an inoculation: no mundane conspiracy could possibly capture that pulse of inside knowledge of a mystic firmament of lies and outlandish goals canopied above our earth-chained heads.

But here I am again, though much less amused and more fearful.

I think I read ten different reporting and opinion pieces today on the topic of Marjorie Taylor Greene, the shock-curiosity of the day who amplified QAnon, Jewish space lasers, political assassination fantasies, and likely a range of yet-to-be-discovered subjects of scorn and ridicule. Most analysts agree that such fantastical and angry ideas are methods for manipulating gullible people. They are tools for the acquisition of power over others.

The whole project feels like an alternative reality so late in America’s evolution, like we’ve transitioned to a Counter-Earth or Bizarro Htrae or Nabakov’s AntiTerra.… Read the rest

Type 2 Modular Cognitive Responsibility for a New Year

Brain on QI’m rebooting a startup that I had set aside a year ago. I’ve had some recent research and development advances that make it again seem worth pursuing. Specifically, the improved approach uses a deep learning decision-making filter of sorts to select among natural language generators based on characteristics of the interlocutor’s queries. The channeling to the best generator uses word and phrase cues, while the generators themselves are a novel deep learning framework that integrates ontologies about specific domain areas or motives of the chatbot. Some of the response systems involve more training than others. They are deeper and have subtle goals in responding to the query. Others are less nuanced and just engage in non-performative casual speech.

In social and cognitive psychology there is some recent research that bears a resemblance to this and also is related to contemporary politics and society. Well, cognitive modularity at the simplest is one area of similarity. But within the scope of that is the Type 1/Type 2 distinction, or “fast” versus “slow” thinking. In this “dual process” framework decision-making may be guided by intuitive Type 1 thinking that relates to more primitive, older evolutionary modules of the mind. Type 1 evolved to help solve survival dilemmas that require quick resolution. But inferential reasoning developed more slowly and apparently fairly late for us, with the impact of modern education strengthening the ability of these Type 2 decision processes to override the intuitive Type 1 decisions.

These insights have been applied in remarkably interesting ways in trying to understand political ideologies, moral choices, and even religious identity. For instance, there is some evidence that conservative political leanings correlates more with Type 1 processes.… Read the rest

One Shot, Few Shot, Radical Shot

Exunoplura is back up after a sad excursion through the challenges of hosting providers. To be blunt, they mostly suck. Between systems that just don’t work right (SSL certificate provisioning in this case) and bad to counterproductive support experiences, it’s enough to make one want to host it oneself. But hosting is mostly, as they say of war, long boring periods punctuated by moments of terror as things go frustratingly sideways. But we are back up again after two hosting provider side-trips!

Honestly, I’d like to see an AI agent effectively navigate through these technological challenges. Where even human performance is fleeting and imperfect, the notion that an AI could learn how to deal with the uncertain corners of the process strikes me as currently unthinkable. But there are some interesting recent developments worth noting and discussing in the journey towards what is named “general AI” or a framework that is as flexible as people can be, rather than narrowly tied to a specific task like visually inspecting welds or answering a few questions about weather, music, and so forth.

First, there is the work by the OpenAI folks on massive language models being tested against one-shot or few-shot learning problems. In each of these learning problems, the number of presentations of the training data cases is limited, rather than presenting huge numbers of exemplars and “fine tuning” the response of the model. What is a language model? Well, it varies across different approaches, but typically is a weighted context of words of varying length, with the weights reflecting the probabilities of those words in those contexts over a massive collection of text corpora. For the OpenAI model, GPT-3, the total number of parameters (words/contexts and their counts) is an astonishing 175 billion using 45 Tb of text to train the model.… Read the rest

The Abnormal Normal

Another day, another COVID-19 conspiracy theory making the rounds. First there was the Chinese bioweapons idea, then the 5G radiation theory that led to tower vandalism, and now the Plandemic video. Washington Post covers the latter while complaining that tech companies are incompetently ineffectual in stopping the spread of these mind viruses that accompany the biological ones. Meanwhile, a scientist who appears in the video is reviewed and debunked in AAAS Science based on materials she provided them. I’m still interested in these “sequences” in the Pacific Ocean. I’ve spent some time in there and may need to again.

The WaPo article ends with a suggestion that we all need to be more skeptical of dumb shit, though I’m guessing that that message will probably not reach the majority of believers or propagators of Plandemic-style conspiracy thinking. So it goes with all the other magical nonsense that percolates through our ordinary lives, confined as they are to only flights of fancy and hopeful aspirations for a better world.

Broadly, though, it does appear that susceptibility to conspiracy theories correlates with certain mental traits that linger at the edge of mental illnesses. Evita March and Jordan Springer got 230 mostly undergraduate students to answer online questionnaires that polled them on mental traits of schizotypy, Machiavellianism, trait narcissism, and trait psychopathy. They also evaluated their belief in odd/magical ideas. Their paper, Belief in conspiracy theories: The predictive role of schizotypy, Machiavellianism, and primary psychopathy, shows significant correlations with belief in conspiracies. Interestingly, they suggest that the urge to manipulate others in Machiavellianism and psychopathy may, in turn, lead to an innate fear of being manipulated oneself.

Mental illness and certain psychological traits have always been a bit of an evolutionary mystery.… Read the rest

The Retiring Mind, Part V: Listening and Ground Truth

Human hearing is limited in the range of frequencies that we can discern. Generally, at the high end, that limit is around 20kHz, which is a very high pitch indeed. But, as we age, our high frequency perception reduces as well, until we may very well have difficulty hearing 8kHz or understanding human utterances in old age. You can test your own approximate limits with a simple YouTube video that raises pitches quickly up through the spectrum. I’m capping out at just north of 13.5kHz using a cheap speaker attached to my monitor, and with normal but quiet ambient background noise.

The original design of the Compact Disc by Phillips and Sony used the 20kHz limit as guidance for the encoding of the digital information on the disks. Specifically, the input analog waveform was sampled at a resolution of 16 bits 44.1kHz, which gives a maximum volume range of 2^16 (96dB) and supports the Nyquist sampling theorem that requires double the maximum frequency of the input stream in order to reconstruct that stream.

And CDs were very good, exceeding the capabilities of vinyl or cassettes, and approaching the best magnetic tape capabilities of the time. They also had some interesting side-effects in terms of mastering by freeing bass frequencies that had to be shifted towards the central channel on vinyl in order to avoid shortening recordings unduly because of the larger groove sizes needed to render low frequencies.

But now, with streaming, we can increase our resolution still further. Qobuz and Tidal offer Hi-Res audio formats that can range up to 24 bit resolution at 192kHz sample rates. Tidal also promotes MQA (Master Quality Authenticated) format that may use lossy compression but preserves aspects of the original master recording.… Read the rest

Forever Uncanny

Quanta has a fair round up of recent advances in deep learning. Most interesting is the recent performance on natural language understanding tests that are close to or exceed mean human performance. Inevitably, John Searle’s Chinese Room argument is brought up, though the author of the Quanta article suggests that inferring the Chinese translational rule book from the data itself is slightly different from the original thought experiment. In the Chinese Room there is a person who knows no Chinese but has a collection of translational reference books. She receives texts through a slot and dutifully looks up the translation of the text and passes out the result. “Is this intelligence?” is the question and it serves as a challenge to the Strong AI hypothesis. With statistical machine translation methods (and their alternative mechanistic implementation, deep learning), the rule books have been inferred by looking at translated texts (“parallel” texts as we say in the field). By looking at a large enough corpus of parallel texts, greater coverage of translated variants is achieved as well as some inference of pragmatic issues in translation and corner cases.

As a practical matter, it should be noted that modern, professional translators often use translation memory systems that contain idiomatic—or just challenging—phrases that they can reference when translating new texts. The understanding resides in the original translator’s head, we suppose, and in the correct application of the rule to the new text by checking for applicability according to, well, some other criteria that the translator brings to bear on the task.

In the General Language Understand Evaluation (GLUE) tests described in the Quanta article, the systems are inferring how to answer Wh-style queries (who, what, where, when, and how) as well as identify similar texts.… Read the rest

Bereitschaftspotential and the Rehabilitation of Free Will

The question of whether we, as people, have free will or not is both abstract and occasionally deeply relevant. We certainly act as if we have something like libertarian free will, and we have built entire systems of justice around this idea, where people are responsible for choices they make that result in harms to others. But that may be somewhat illusory for several reasons. First, if we take a hard deterministic view of the universe as a clockwork-like collection of physical interactions, our wills are just a mindless outcome of a calculation of sorts, driven by a wetware calculator with a state completely determined by molecular history. Second, there has been, until very recently, some experimental evidence that our decision-making occurs before we achieve a conscious realization of the decision itself.

But this latter claim appears to be without merit, as reported in this Atlantic article. Instead, what was previously believed to be signals of brain activity that were related to choice (Bereitschaftspotential) may just be associated with general waves of neural activity. The new experimental evidence puts the timing of action in line with conscious awareness of the decision. More experimental work is needed—as always—but the tentative result suggests a more tightly coupled pairing of conscious awareness with decision making.

Indeed, the results of this newer experimental result gets closer to my suggested model of how modular systems combined with perceptual and environmental uncertainty can combine to produce what is effectively free will (or at least a functional model for a compatibilist position). Jettisoning the Chaitin-Kolmogorov complexity part of that argument and just focusing on the minimal requirements for decision making in the face of uncertainty, we know we need a thresholding apparatus that fires various responses given a multivariate statistical topology.… Read the rest

Bullshit, Metaphors, and Political Precision

Given this natural condition of uncertainty in the meaning of words, and their critical role in communication, to say the least, we can certainly expect that as we move away from the sciences towards other areas of human endeavor we have even greater vagueness in trying to express complex ideas. Politics is an easy example. America’s current American president is a babbling bullshitter, to use the explanatory framework of the essay, On Bullshit, and he is easy to characterize as an idiot, like when he conflates Western liberalism with something going on exclusively in modern California.

In this particular case, we have to track down what “liberal” means and meant at various times, then try to suss out how that meaning is working today. At one time, the term was simply expressive of freedom with minimal government interference. Libertarians still carry a version of that meaning forward, but liberalism also came to mean something akin to a political focus on government spending to right perceived economic and social disparities (to achieve “freedom from want and despair,” via FDR). And then it began to be used as a pejorative related to that same focus.

As linguist John McWhorter points out, abstract ideas—and perhaps especially political ones—are so freighted with their pragmatic and historical background that the best we can say is that we are actively working out what a given term means. McWhorter suggests that older terms like “socialist” are impossible to put to work effectively; a newer term like “progressive” is more desirable because it carries less baggage.

An even stronger case is made by George Lakoff where he claims central metaphors that look something like Freudian abstractions govern political perspectives.… Read the rest