TuringTest

Chile was building early some feedback-based information systems like what modern corporations use. They may have worked or failed, but we will never know because the US government indirectly promoted and financed a military coup that installed an actual murderer totalitarian regime.

https://en.wikipedia.org/wiki/United_States_intervention_in_...

Just wanted to point out that the variation of sound is not random, but with the sound tied to the letter. Also Spain has something that English does not, which is an official organisation studying the language and making official rules recognising how it's used at different contexts and zones.

Some regions still make that difference, at least senior people do. I had a language teacher from old Castille who sound her spelling of v and b very clearly. She made the distinction between 'll' and 'y' clear as well, with 'll' sounding as a modified 'l', a cross between 'li' and 'ye'.

I thought the whole point of writing programs in C was originally being able to write portable software (and a portable OS) that could be executed at different machines? It only became specialized in giving machine-specific instructions when other languages took over.

Fighting for civil rights often makes you look like a prick, because you keep laser-focused on your goal and need to counter all the reasonable-sounding objections of people who were following their daily routines before this ball-breaker came along; but it is nevertheless necessary.

Contrary to Hollywood films, people don't stamp on other people's rights because they have some inner impulse to do evil, but because injustices are ingrained in the common way to do things, and fixing then implies to deviate from those routines; that's why it's so hard to change them.

That's the real meaning of the sentence "for evil to triumph, all it takes is for good people to do nothing". The movie script of a hero taken the matter in their hands and saving the world with heavy guns is but a fantasy

"Purpose" is something we project onto the world, not something that nature possesses.

A summary of everything that's going on in a scientific field may allow reviewers to find out which processes are solid, which ones are grounded on faulty premises or bad data, and which ones merit further research. Doing the same without the AI summarisations could take decades of review work.

https://lawsofux.com/aesthetic-usability-effect/

https://en.wikipedia.org/wiki/Emotional_Design

The moral of the story, if there's one, is that constraints may be useful if they reflect common design rules and the system is built around those cases with high level abstractions, but simply allowing any possible low-level constraint may be too complex for the case of designing GUIs.

On the contrary, every new way to prove a known theorem has the potential to be applicable in other areas of the same or related fields, extending the mathematicians' toolset with new instruments. These new methods often serve as a seed for new discoveries.

It's better to think of this models as 'generating' chains of relevant words, where 'relevant' is defined by similarity of those areas of knowledge on which it has been trained, and which are "activated" as close to the topics in the prompt. Which is not at all dissimilar to how humans learn about a new topic, btw.

This way, by "activating" concepts of areas of knowledge and finding words that are more likely than others to fit those concepts, the model is able to create texts following the constraints you instruct it with - such as poems that rhyme, or critical analysis of scientific articles.

The most important point to be aware of is that this creation model is completely different to how automatic reasoning models create content, which is by having a formal representation of a knowledge domain and creating logical inferences that can be mathematically proven correct within the model. A reasoning model cannot lie, but it cannot create content beyond the logical implications of its premises; its quite the opposite of what language models do.

What I'm warning is against thinking of them as independent agents with their own minds, because they don't work like that at all, so you'd be anthropomorphising them.

These models certainly have a compilation of knowledge, but it is statistical knowledge - in the same way as a book of logarithms has lots of mathematical knowledge, but you wouldn't say that the book 'knows logarithms'. The compilation contains statistical 'truths' about the topics on which it has been trained; and contrary to a written book, that knowledge can be used operationally to build new information.

Yet that static knowledge does not reach the point of having a will of its own; there is nothing in the content generation system that makes it take decisions or establish its own objectives from its statistical tables of compiled knowledge.

They are excellent imitators, being able to clone the style and contents of any subject or source you ask for. When you prompt them, they will uncritically generate a text that combines the relevant topics in creative ways, without the least understanding of their meaning.

Their original training causes them to memorize lots of concepts, both high and low level, so they can apply them while generating new content. But they have no reception or self-assessment of what they are creating.

For a start, GPT-4 doesn't include in its generation the current state of its internal knowledge used so far; any text built can only use at most the few words already generated in the current session as a kind of short-term memory.

Biological brains OTOH have a rhythm with feedback mechanisms which adapt to the situation where they're doing the thinking.

It is entirely possible to build mechanical thinking in organic material (think Turing machines built on growing tissue), and it could also be possible to build complex self-referential processes simulated on electronic hardware, of the kind high-level brains do, with their rhythms of alfa and beta waves.

Also you're missing the obvious solution, using a long-term support system with a robust audit process and well timed security releases. Author's setup will NOT be safer than that.

And it's equally an incomplete assessment based on the same "This is fine" attitude. The computer could be participating in a botnet dedicated to bring down the network in hospitals or steal their sensitive medical data, and the author would never know.

As for visualizations, you may want to take a look at mathematical relations [1]. Each cell is a variable with the alphabet as its domain (which is pruned as you guess correct values), and the words are 5-arity relations between rows and columns. You may represent the local relations surrounding the cell pointed by the mouse.

[1] https://en.wikipedia.org/wiki/Relation_(mathematics)#Represe...

In the same way that wrong results in a search engine will not be solved or mitigated in its entirety, yes. These new generative AIs are big search engines that bring results from several content points and combine them into a single, plausible stream of words.

In that case, this current batch is not "advanced AI". It is a big autocomplete panel which retrieves content already present in its training corpus, and presents it in new shapes; it is not ready to define new goals with a purpose and follow them to completion.