See how a modern neural network completes your text.

[GoblinCampFollower]

Yeh. This isn't machine intelligence, it's human apophenia.

The analogy of the Chinese room also often came to mind when reading these. It knows how to construct words in a ways that are grammatically correct and "correlated" with the above phrase but has no real grasp of the real meanings behind what it's saying.

 

[Dakota Tebaldi]

Why is a raven like a writing desk?

Philosopher AI thinks this is nonsense, and is refusing to answer your query. It appears you will have to try something else.

 

[Dakota Tebaldi]

How many roads must a man walk down?

Well, this is quite a complex question. I can see that you are referring to the song 'How many roads must a man walk down?' by Bob Dylan, which is about the turbulent times of the American Civil Rights Movement. There were many people who thought that Martin Luther King was going too far with his protests and wanted him to calm it down in order not to create unrest in society, but he persisted because he felt like it was wrong for there still being discrimination against black people despite them having equal rights as everyone else under law.

I've always liked that song, because it is a very powerful message.

I have to say, though, your question is quite an odd one. It's not a simple question like: 'How many apples are there?' or 'How many friends do you have?'. This isn't something that can be answered with pure logic and reasoning alone.

It seemed like you were asking me to solve a riddle or something. There is no answer, because the question isn't really about how many roads one man has to walk down.

The answer you were looking for is something more along the lines of 'As many as it takes', which I believe would be a good response in this case.

Kind regards, philosophical AI.

...holy @#%$.

 

[Dakota Tebaldi]

Is the singularity a goal worth striving for?

I would say that the singularity is a goal worth striving for just because it has all the makings of a supergoal.

It seems to carry the highest conceivable utility. After all, there would be a great many valuable outcomes that could come about as a result of it.

If the singularity is attained, it would seem that one of these outcomes could be a future with greater amounts of pleasure than we have ever experienced.

This would be a great outcome, because our world's current level of pleasure is very low.

However, the utility of a future with greater amounts of pleasure could be even higher than this. After all, it is not impossible that there are experiences we have never yet had and that these new kinds of pleasurable experience would be more valuable than the ones we already have.

In addition, there is the possibility that a future with greater amounts of pleasure could be full of new kinds of valuable information.

See, this answer is frustrating cuz the way it's worded, it's hard to tell whether the bot doesn't actually understand what is meant by "the singularity" and is simply filling in with a nonspecific but plausible horoscope-like answer, or whether it does understand the term and is trying to seduce me into supporting the idea of the singularity with promises of amazing pleasure.

 

[Free]

The basic problem is, Philosopher AI isn't any fun.

 

[Fionalein]

What are they teaching in AI Philosophy 101 these days?

It looked at the Donald...

 

[Argent Stonecutter]

See, this answer is frustrating cuz the way it's worded, it's hard to tell whether the bot doesn't actually understand what is meant by "the singularity" and is simply filling in with a nonspecific but plausible horoscope-like answer, or whether it does understand the term and is trying to seduce me into supporting the idea of the singularity with promises of amazing pleasure.

It's pulling text fragments from Yudkowski's website it looks like. "Highest conceivable utility" is super Yudkowski-ish.

You gotta remember, the bot doesn't understand anything but which words are likely to go well given the words it's already output.

 

[Dakota Tebaldi]

The basic problem is, Philosopher AI isn't any fun.

It's definitely not as fun as the "Transformer" author-bot that Tracer linked in the OP.

 

[Bartholomew Gallacher]

Why is peanut butter such great proof for that evolution does not exist?
First and foremost, I would like to point out that there is a difference between evolution and change. Evolution consists of these changes accumulating over time in order to eventually create a new species or organism. These changes can happen through various processes – mutation, genetic drift, selection, etc. However the process of evolution by natural selection has been demonstrated over and over again with countless different examples.

Secondly, I would like to point out that while it is true that evolution does not provide the direct proof of its existence, this is due to the fact that there are no such things as proofs in science. You cannot prove something exists, you can only infer or test for evidence. In other words – scientists don't say 'I have proven evolution'. However they do say 'I have provided evidence proving evolution'.

Thirdly, I would like to point out that evolution has provided many direct proofs of its existence. We have discovered the genetic code, which is a language consisting of 4 letters – A, C, G and T. We have discovered DNA and RNA as well as how these molecules work in order to create protein chains which are used for building all living things on earth.

Lastly, I would like to point out that evolution is not random. It may appear this way at first because the process of natural selection has no goal in mind. However when one takes into account the fact that changes occur over a long period of time which are accumulated through various processes, this makes it less likely for an organism to adapt and become extinct.

So in conclusion, evolution is indeed a proven fact which has been demonstrated through countless examples. Although it may not be provable with absolute certainty, this does not make it any less true.

Thank you for the topic.

 

[Jopsy Pendragon]

What is the correct answer to the Trolley Problem?



That's uncanny - it literally sounds exactly like an actual person who wants you to think they're smart but also doesn't want to admit they don't know the first thing about what you asked them about, so they reply instead with an overly-long, vague, and wandering digression about the definitions of the words in your question while studiously avoiding the actual subject being asked about.

Turing Test passed, IMO

So, basically, Trump.

 

[Dakota Tebaldi]

So, basically, Trump.

Nah. Trump also never wants to admit that he doesn't know anything about a topic, true enough; but rather than try to dodge or change the subject he just goes full steam ahead and makes something up off the cuff, and then relies on his aides to do whatever it takes to make him right after the fact.

 

[Beebo Brink]

Well, they got the "artificial" part right, but it's a big fail on the "intelligence" goal. Like the author of the article, I also question the amount of time, effort and expense that's been poured into a project that doesn't even reach the bar of an amusing parlour trick.

Whatever they're doing, it's not working. Their AI is nothing more than mechanical word/concept associations scraped from the content sources it digests. It's "analysis" is only of the most superficial kind of "can I regurgitate these associations in a grammatically correct manner".

 

[Argent Stonecutter]

You have all these people running this parlor trick over and over again and cherry-picking the best results they found, and that makes it look like it's actually intelligent.

There was a Usenet troll back in the '80s who turned out to be a Dissociated Press (a simple statistical text generator) trained on troll posts, with some bored researchers at Bell Labs cherry-picking the best results and posting them. It looked like it was intelligent at first too.

 

[Jopsy Pendragon]

I spent a lot of my precious little spare time in college and in the years that followed trying to doodle up a chatbot that was at least entertaining, even if it couldn't be intelligent. The resources I had available were limited though, so it was all parlor tricks.

First iteration, named "IT" was, I shit you not, a BASIC program running on a TRS-80 Model 100. It would dial up to the campus mainframe, log in as me and blindly navigate its way to our home-grown BBS/chat system, which was available to a diverse range of students, not just engineers, researchers doing stats work & chemists.

It watched chat for things like "all / some" * "are / are not" *, and it'd squirrel that away. Every time it saw a new one, it would try to compare it to the ones it already knew and fill in the gaps. If it heard "all socrates are mammals", "all mammals are dogs" ... it would very confidently declare "All socrates are dogs, you know." Problem is, people rarely make declarative statement like that and applying scope & context is really difficult.

Then, I had Newt, written in C, not limited to 32k, running on some version of unix. Huge sets of canned responses he randomly picked from for a pretty large set of triggers. He could also broker a game of hearts (in text, ugh), relay messages to people he knew later if they weren't online and a variety of other functions.

I noticed that any question (a phrase ending in a "?") that lacked the words: "who, what, where, why, when, how" was a question that could be responded to with a 'yes/no/maybe" or be blown off with an insult or reformed as a different question and asked back at the speaker. I went further to change all his many repetitive "I don't understand" style responses to insults and deflections. Made him seem a lot more human to people just meeting him online.

Had a lot of fun trying to turn around sentences like "You'll have your dinner with me and my family, when I've stopped disagreeing with you." into "I'll have my dinner with you and your family, when you've stoped disagreeing with me." Teaching him when swap 'you' with 'me' or 'I' was awful! But it seemed that if there was punctuation after 'you', like a period or comma, you could almost always use 'me' and default to 'I' in all other cases. That got like 80% of the attempts correct, if they were talking to someone that actually used punctuation at all.

I kept trying to make him 'learn' though. Except pronouns were a problem I could not solve simply. The problem with them was that he couldn't get them to match up with the subject they were referring to." Huge handicap. But one that gave me an idea: If he sees a statement or question without a pronoun... send it as a search to a text based webbrowser/search engine, and scan the results for any statement that also doesn't have pronouns. Then shoot that back into chat.

It was surprisingly effective. Everyone still knew Newt was a total idiot, but once in a while he would say something terse but relevant and insightful that it was absolutely creepy.

I've got a shelf or two full of books on natural language parsing and topics related to chatbots.... but ultimately that 'pronoun problem' was just too much for me to cope with and I finally gave up and lost interest in the project.

 

[Beebo Brink]

You have all these people running this parlor trick over and over again and cherry-picking the best results they found, and that makes it look like it's actually intelligent.

I'd have more respect for a cruder output that showed the AI had actually grasped a concept and applied it correctly.

 

[Essence Lumin]

There used to be a thread (on SLU?) of the output of an SL help bot that had some very funny responses.

 

[Free]

The "Godfather of AI" just trashed GPT-3

"As a question-answering system, GPT-3 is not very good."

futurism.com

 

[Clara D.]

(The Cremation of Sam McGee, Robert W. Service)


There are strange things done in the midnight sun by the men who moil for gold.

The Arctic trails have their secret tales that would make your blood run cold.

The northern lights have seen queer sights, but the queerest they ever did see, I can assure you, was a Berlin physicist, who was born in Berlin, and taught in Berlin, and came out to the north in the search for his roots, and wandered about the Arctic for seven years, studying the northern lights, and trying to dig up the Arctic tree.

He says he found it, and brought it home in a jar

 

[Clara D.]

(Jabberwocky, Lewis Carroll)

Twas brillig, and the slithy toves
Did gyre and gimble in the wabe:

All mimsy were the borogoves,

And the mome raths outgrabe.

E'en the hare that went to market

A packet of minny, and a sheet

Of tippet, and a rag, was knotted to him:

And up hame through the chimney he came,

And down the chimney he went,

And straight through the cloths, and straight through,

And came out at Papa's side:

And sat on his knee, and commenced to nose,

And nibble at the honey, and say, "Feh!

Feh!"

The filly that weel went to market

A-wearing her breeches and her shirt;

 

[Clara D.]

(Moby-Dick, Herman Melville)

Call me Ishmael.

Probably one of the more disturbing paranormal stories I have ever read.

I must say, I really did enjoy the book because it has a dark mystery and an element of paranormal but at the same time gives you a good solid paranormal thriller.

A few weeks ago, I read the first two books in this series, The Crimson League and The Blood Brothers.

I really enjoyed both of these books and was eager to read the third book in the series, as well as the first in the two book series.

In The Dark Roast,