I would love a social experiment. If you had the time or resources, allow two different comment feeds and let the market decide. One feed would be the moderated anonymous feed. The second would be registered non-anonymous. Anyone can see and comment in the anonymous feed. Only registered users could comment in the non-anonymous feed. Where would users go for thoughtful analysis?
There would still be a choice to allow everyone to read the non-anonymous feed or maybe part of the experiment would be to restrict the reading to registered users.
It would be interesting to see how the market (the public users) would self select. Would the non-anonymous feed be more relevant? Does an anonymous feed actually encourage more free thinking or is that a myth. (Yes, this would be subjective although incivility is fairly objective)
Your change is sad. Maybe it doesn’t need to feel like a defeat. Your past 20 years could be considered an exceptional success. Let’s hope some of this uncivil current reality is a short blip.
]]>Re : End of a freedom.
With regards,
“This is a security failure. We have been defeated by those that wish the internet and even humanity ill.”
I saw it coming “in a way” in what was happening years ago. It started with certain people making personal not objective criticism against an individual as a singular agenda. But never making anything approaching a societal contribution that was constructive, relevant, or original.
Their intent back then was to damage reputations in the eyes of casual readers[0]. This inturn encouraged others to “go scalping” for what I assume was petty bragging rights in small cliques of mostly irrelevant interest to the greater majority.
For a while such behaviour was popular and mostly based on fake comments that got drummed up by others “looking for topics” to write up or gain their 15 seconds of fame. It did a lot of harm for a while untill others started to realise it was possible for those of ill intent to create smoke and noise without there actually being a fire. But in the process they did burn a lot of people.
The supposed right of “free speech” is a precious gift to be used wisely. Like all tools it is a double edged sword that can be used for good or bad. Misuse to polish a personal ego or gain fame or notoriety via a baseless agenda –which is what all to many have tried to use it for, often infamously so– damages free speech with each such use in the eyes of society.
Which is why over the years Judges have decided there are limits to “free speech” which is why even school kids used to get told “Shouting fire in a crowded auditorium” is not free speech. In part because of the considerable harm it does to innocent people.
The principle behind this cautionary behaviour is much as it is with early professions of “First do no harm”. Unfortunately the lack of dealing with such bad behaviours led upto the point it “became a thing” of Social Media one name of which was “Cancel Culture”[1].
Whilst originally providing an outlet for those with a genuine need to “make public” which Free Speech alows, it has quickly become not just tainted, but weaponised for politics and personal agendas as will no doubt become abundantly clear over the next few months.
Whilst “political idiots” / “talking heads” espouse the power of AI to “moderate” they clearly do not understand the fact that the current LLM and ML systems really can not do such a job currently, and nor are they ever likely to. Because the LLMs are not in any way “intelligent” nor are they in any way “predictive” just grossly laggingly “responsive” at inordinate cost/harm.
But still worse the cost of keeping these “Deep” “Digital Neural Networks”(DNNs) even remotely upto date via ML is inordinate resource hungry almost beyond measure. Hence the use of the less expensive “sweat shop labour” to add layer after layer of filtering, which actually makes the resource usage issues worse.
Whilst it was clear to me that the “agenda / fame” nonsense was going to rise and I gave repeated warning, it turns out others were thinking in this area both more intently and academically. One such is a fellow at the Berkman Center, Dr. Aaron Shaw[3] who back over a decade ago in 2012 had a paper published,
“Centralized and Decentralized Gatekeeping in an Open Online Collective”
Even though based on observations of behaviours back in 2008 it is still an interesting read on the non technology effectively social aspects of the subject of moderation. From which it can be seen why even other forms of AI that are not the current LLM or AI systems or those likely to spawn from them will not function as effective moderation either.
Which should not be at all surprising, as the past several millennium of human history shows trying to “social censor” against society, no matter how draconian the enforcement, in the end always fails to human ingenuity be it by the like of in channel satire, innuendo, or more subtle means. Or the simple fact control on one channel is insufficient as like ink on a wet page information bleeds across from line to line and even from page to page.
But all that aside learning is a two way process, because multiple view points give rise to thoughts, considerations and questions that might not have otherwise arisen. Thus this blog has been very much a collective effort. And for those that have read our hosts writings can see the influence this blog has had in turn on him, and other well know academics. Some of whom, Ross J. Anderson and Nicholas Weaver were regular contributors with others like Moti Yung reading and occasionally commenting.
The thing is, as was once noted, the simple and innocent question of a child if considered can give rise to a lot of deep thinking. Such as “Why is the sky blue?” or “Why are clouds white?” or the all time favourite subject of many children “What is a rainbow?”. If we truly know the answers then we can tell or show children the answers in ways that will encourage them to understand more themselves and so encourage others. Whilst the world appears to be magic, that is superficial and a deep understanding actually does not rob the world or individual of charm, it enables them to go on and create magic for following generations.
Sadly the “new policy” of the likes of Google is not to make the world available to all any longer. They have decided that you will be led astray by their choices, unless you know enough to be able to push through their bias to what is out there.
What ever their supposed “corporate reasoning” we can make an assumption it has actually to do with that triad of “Money, Politics and Power” that act as the tools of control. It actually brings forth all the human evils we associate with Kings and Bishops and their unsavoury efforts to keep people in oppression and subjugation, not for the good of society no matter how much they might claim, but for the evils that they so revel in.
This blog has acted as a place of knowledge where the triad of tools got ignored, and the control others sought not just questioned but challenged and railed against.
[0] And now of course that this blog has been scraped into current AI LLM systems apparently that ill intent has spread out in a way “search engines” could/did not previously do.
[1] In her 2022 essay / e-Book “Cancel Culture: A Critical Analysis”,
https://link.springer.com/book/10.1007/978-3-030-97374-2
Academic author Eve Ng[2] pointed out in it, her definitions of the terms “Cancelling”, and “Cancel Culture” that both are the practice of nullifying or cancelling someone or something by a process of “speaking out” or “shaming”. That is in someone’s eyes they have done “bad” in some way, hence the speaking out about them. Be they an individual, a group of individuals, an organisation, commercial or other brand, or as we are currently seeing even entire nations. By so called “Public Shaming” in what is often not a fair forum, and seeking to dictate the surrounding commentary about the alleged bad / wrong doing.
In short what first started out as a “social good” has quickly become a weapon of politics and in many places a “social bad”.
You can get a feel for the drivers behind the essayt e-Book with,
https://m.youtube.com/watch?v=95JuPhRjDn4
So to be honest I don’t recommend it as a general read because it becomes somewhat tainted and biased by what reads like cognitive bias or a cherry picking process.
[2] Associate Professor Dr. Eve Ng was at the time of writing her e-Book / essay[1], based in the School of Media Arts and Studies at OHIO University USA. As well as the associated Women, Gender, and Sexuality Studies programs with a focus on internationalisation of minority groups in small nations via online means to find/make creative spaces.
She has the problem that there is a computer security suite “EVE-NG” that coincidentally shares her name, which makes searches via the “main engines” a bit difficult.
[3] Professor Aaron Shaw based at the Department of Communication Studies at Northwestern University and also a fellow at the Berkman Center for Internet and Society at Harvard University.
As part of his research into the dynamics of large online communities and the participation, mobilization, and organization of individuals and groups surrounding them back in 2008 he published in 2012 the paper,
“Centralized and Decentralized Gatekeeping in an Open Online Collective”
Which details some of the issues of what we would now call “Comment Moderation”.
]]>I have learned very very much from the comments on the Squid section. So first:
Secondly, I feel this is a terrible defeat. I have come to realize that there is a huge value in long running old communities of discussion. There is a tacit knowledge that builds up and a value in comments that come from people like Clive who have built a reputation over years. This value is not just in Clive alone, but also the knowledge that if there’s something missing it will be called out and Clive will give a clear answer.
The anonymity here has allowed interesting and important comments that would never have happened in a more controlled forum. That was also critical to the sum of the value of the comments. The toxicity was unfortunately inevitable.
There are real conscious reasons to destroy that community. Many people will benefit from that. Few of those people are good or have the interests of humanity in mind.
This is a security failure. We have been defeated by those that wish the internet and even humanity ill. I hope we all acknowledge that. I would propose that everyone considers whether there is another place that this community could agree to move to. Without that, the same will not grow up and a little pearl of wisdom in the world will be lost forever.
]]>Head over to hugging-face (or find someone to help you) and grab the latest model, fire it up on an old gaming PC with a decent graphics card, and you should be able to quickly tell it how to score comments (is it on-topic? is it hate? is it spam? …) then leave it running with an agent to take action based on those scores.
Remember to practice threat-deception – do not let people who comment know that you erased their comment, or else they’ll just come back and keep trying until their garbage shows up.
]]>On the other hand, glad to see you taking action against toxicity and hateful speech!
]]>Part of this discussion is based on “what do we expect as output” from such systems.
Indeed. Examples for DNNs are, the words that were said, or the contents of an image, or the identity of a speaker, or a face, or a fingerprint. But it can also be the answer to a question, an image fitting a description, a summary of a text, or the sentiment of a comment.
In those cases there is no well defined “correct” answer, only a “likely” or “most likely” answer. That is, any answers have to be rated using some kind of (bayesian) statistics.
such that the query to return Fruit As Apple results in Red Apple, Green Apple and Pineapple.
This depends entirely on how the system is trained and what input data is used. For instance, LLMs only look at the context tokens are used in. If the input data show pineapples appear in different contexts, eg, pizzas, than apples, eg, in pies, then they will not appear easily in the same answers.
If you want a system that takes note of the biological taxonomy of plants, then the training data abs procedures should support that.
Red is selected over Purple because the averages on Red+Apple are higher and there aren’t many use-cases for Purple+Apple whereas Purple+Cow is a common expression.
You are now generalizing simple Markov chains to DNNs. That does not work. Markov chains are unable to capture the syntax or semantics of human language. DNNs can capture the syntax of human language. [1]
[1] Markov chains capture regular grammars, human languages have context sensitive grammars. In computational terms, these grammars are recognized by, respectively, finite state automata and linear bound automata.
‘https://www.google.com/amp/s/www.geeksforgeeks.org/chomsky-hierarchy-in-theory-of-computation/amp/
You might find this of interest,
]]>re: AI building non functional systems
Part of this discussion is based on “what do we expect as output” from such systems.
In basic computer systems we expect “correct behavior” as defined by the PRD and System Designers; such that if the “correct behavior” expected is AnswerA that is the result we attempt to achieve.
In AI systems, correct answers are not required and even penalized in some aspects due to how the systems determine what is a correct output; such that the query to return Fruit As Apple results in Red Apple, Green Apple and Pineapple.
In this later aspect, the AI correct output of a system is totally acceptable and produces Pineapple in the output results. It is not the answer humans accept, although they might find it hilarious when William Tell shoots a Pineapple off of the head of his son.
All (afaik) human languages have grammar and organization. These are not the same between languages and ethnicities however, there are required word-order patterns. In English we have adverbs and adjectives, which are represented by the Token Set in AI for their “inference” substitutions. A data-token-set designated as an adverb will fit into any adverb position in an English sentence.
In AI it doesn’t matter what word it picks as long as it fits into the correct slant line. There is a statistical component that indicates the rarity of the word and a weighted average used to pick a more likely one (Sherlock Holmes The Adventure of the Dancing Men). Red is selected over Purple because the averages on Red+Apple are higher and there aren’t many use-cases for Purple+Apple whereas Purple+Cow is a common expression.
This is the limitation and eventually the failure of AI as currently designed. The difference between what “appears” to be OK and what is “nonsense” or worse “dangerous” outputs.
RL Difference tl;dr
In most book marts there are lots of books on horseback riding. Some are picture books for children enchanted by the mere thought of horses and ponies and some are thick scholarly tomes on the topic which take decades at best to absorb their meaning.
The main reason books fail past more than the basic information is that horse riding is something that cannot be learned from a book. It is by direct experience that one learns to ride. Scholarly tomes are attempts at describing intricate movements and muscle tensions in word format but they cannot be realized except by enacting them on a horse.
When people asked me to teach them to ride and wanted to know how long it would take, I would tell them that in 2-3 days I can teach you to sit on a horse (walk trot canter); but it will take a “lifetime to learn how to ride”.
===
Search Terms
William_Tell
Shooting an apple off one’s child’s head
Sentence diagram
Purple Cow
Lt. Col. D’Endrödy was a member of the Hungarian Olympic Three-Day Event Team, a member of the Hungarian International Show Jumping Team. British national team coach at the 1956 Stockholm Equestrian Olympics they won a gold medal. In 1959, he wrote his book “Give Your Horse a Chance” in England, one of the most outstanding works in the international horse literature.
]]>Re : Incorrect thinking does not build functional systems.
Maybe I can reformulate the principles more clearly.
Neural Nets, artificial or biological, of sufficient dimensions can approximate any computable function. This is called the universal approximation theorem. See:
‘https://en.wikipedia.org/wiki/Universal_approximation_theorem
This has been proven for many different ANN architectures.
A Universal Turing Machine is a computer that can emulate any computational function. The equivalence seems obvious.
This is obviously just an illustration, not a proof, but you can find the relevant proves in my link above and the Wikipedia pages.
]]>