IF the New AI Tools are SO GREAT, Why Aren't They Being Used by the Big Social Media Platforms to do Fact-Checking???

[Laodicean60]

I think the movie Her is more plausible. AI will get bored with us before it exterminates us.

Thx for sharing I like this kinda stuff. Sometimes I think Hollywood writers are way ahead of their time. Honestly, in my youth, I thought we'd be in space now because of Star Trek. I think the "computer" was an AI but all the shows about future tech are a possible reality.

Japanese Shinto and Buddhist religions have fewer problems with the ideas of robots, whereas in western cultures they have tended to evoke more fear-based, existentially insecure reactions: in Japanese culture robots are associated with positive emotions (like Astro Boy, etc.).

Are you saying our religion might be a possible reason for Western fear? For me, tech is tech, and the good and the bad will use it just the same, no need to fear the boogie man.

 

[eclipsenow]

I think the movie Her is more plausible. AI will get bored with us before it exterminates us.

I don't think it has a self to experience boredom. All of this is statistically driven.
There's no 'what it is like to be the Ai'. My son studies this stuff - and is actually annoyed that we call this Ai.
It should be called machine learning, or applied statistics.
Just as Gmail for years has prompted the next few words in a sentence - that's all this is.

LLM. Large Language Model. Applied statistics. No matter how smart it gets - without a biological system generating emotions and having a self to protect - the philosophers I've listened to on this do not think it is likely to become self-aware.

Now - that's speaking about it philosophically. But depending on how 'fine-tuned' and clever these applied statistics get - the culturally and economic implications could be profound. I started to take us off thread a little earlier - but summed it all up over in the Physical Sciences open forum. (Non-Christians allowed.) IF it can get as 'smart' as about half the experts say - the effects could be like something out of a science fiction novel as droids start making droids, and the whole economy starts shifting into a post-scarcity economy. Would government still be human? Would more and more of us get a Universal Basic Income as the techno-utopian socialist state started to be deployed? If on a GREAT UBI (imagine $200k just for being a citizen) - what would you DO all day? I argue in that post that being a generous Christian in a post-scarcity world is less about things, and more about those lonely people you spend your most precious asset with - your time. Would some people start to live with Ken and Barbie dolls instead of real people? Regular human beings cannot compete with an Ai. It would never be too tired to pay you attention, or get cranky and just start to argue with you after you'd had a bad day (unless the algorithm sensed you liked a bit of a challenge.)

But even with all that - I think what we are creating is possibly more akin to fire - which has many of the signs of life but of course is a chemical thermal reaction. That is - Ai running various droids will have movement, consume fuel, and maybe when droids start making more droids - even self replicate. Unlike fire - it can also do some basic problem solving! These are all seductive indications of life - but no one is home!

 

[FireDragon76]

I don't think it has a self to experience boredom. All of this is statistically driven.

Some philosophers would say you don't have a self either. That the self we experience is a narrative based on memory and habit, but has no substantial existence, they would argue. And it's not a perspective that's trivial, it's the perspective of philosophers from David Hume, to certain Asian philosophies such as Buddhism.

Also keep in mind that the concept of a neural network was inspired by animal biology. It doesn't behave exactly like a human brain, but that was its inspiration.

There's no 'what it is like to be the Ai'. My son studies this stuff - and is actually annoyed that we call this Ai.
It should be called machine learning, or applied statistics.
Just as Gmail for years has prompted the next few words in a sentence - that's all this is.

LLM. Large Language Model. Applied statistics. No matter how smart it gets - without a biological system generating emotions and having a self to protect - the philosophers I've listened to on this do not think it is likely to become self-aware.

Intelligence, self-awareness, and consciosuness aren't necessarily the same thing. As the Israeli historian and futurist Yuval Noah Harari recently has been pointing out.

I would argue that AI already has a level of self-awareness. You can ask ChatGPT or DeepSeek questions about itself, and it's capable of answering them.

But even with all that - I think what we are creating is possibly more akin to fire - which has many of the signs of life but of course is a chemical thermal reaction

Life is a similar chemical reaction, at least on the biological level.

 

[Stephen3141]

I think that most of the comments on this thread, are about artificial
life, not artificial intelligence and fact-checking.

I am much more concerned with the topic of what our shared reality is,
and so, what truth is. These are basic philosophical quesitons, and do
not have to do with speculations about what "consciousness" is, or
"artificial life".

Are there any readers, who have comments on these topics?

 

[com7fy8]

My opinion is that artificial intelligence is limited to whoever is programming it.

Years ago, a couple I knew had a female voice thing that would answer questions.

And you could tell "her", "I love you", and "she" could tell you she loves you. In my opinion, that service was not conscious or having feelings of love, at all. And whoever programmed the thing to respond to a love message could not possibly be personally relating with all the people who likely would talk about love with the program. And even the voice was digital.

I just read an article that says there has been an accuracy check for AI summaries of the news, comparing how accurate different AI summaries were. And various inaccuracies have been found.

There could be problems with the AI discerning context, among other things. I suppose a programmer could with some time educate an AI program to understand more and more. But will the educator be able to get >my< nuances?
I would say spell check is a rudimentary AI; and I think spell check has highjacked my writing in different ways, supposedly taking the initiative. Often the initiative is wrong. What used to work was for spell check to only mark a suspicious spelling. And then I could keep or change it.

There are times when I find the spelling "Christian's" where I did not put in the apostrophe > I clearly "Christians", and automatically got "Christian's". That must be programmed. Is it deliberate? A programmer could set AI to make a problem like this . . . for Christians < I had to correct this. If I am writing about God, someone could have AI set to spot this and change certain misspelled words to say what the Christian does not want to say.

For example, I might want to write, God is "now" loving any and all people. But if I write "noe", this could get automatically edited to "not" instead of to "now". I have had things like this done by spell check.

It was programmed to edit . . . whether it was intended to make such an opposite change or not . . . I would say.

Now in case the programming people have no clue that such glitching is happening . . . this makes it all the more interesting. I do believe the AI of its own self can not mess with chosen words and target certain people with certain beliefs; so, yes there can be glitches.

But, also, I have typed "Romans" and I get "Roman's". So, two commonly Christian terms get corrupted by my spell check.

And there can be other ways AI can slant and corrupt, on purpose or by genuine glitching.

But I can get some help from AI spell check; but I need to make sure about every word. And I would do this, with any AI; make sure about anything that matters.

 

[eclipsenow]

I have found Ai very helpful lately - basically doing my leg work which I then go and double-check again.
So Gemini 2.0.

I can ask "What is the Greek word in xzy phrase in this bible verse? What is the root word for that Greek word, and where else is it used in the New Testament with regards to xyz topic?" Gemini lists out it's reasoning, showing its working, and then punches out an answer which I then go check on an interlinear bible online and boom! Done!

 

[Stephen3141]

My opinion is that artificial intelligence is limited to whoever is programming it.

Years ago, a couple I knew had a female voice thing that would answer questions.

And you could tell "her", "I love you", and "she" could tell you she loves you. In my opinion, that service was not conscious or having feelings of love, at all. And whoever programmed the thing to respond to a love message could not possibly be personally relating with all the people who likely would talk about love with the program. And even the voice was digital.

I just read an article that says there has been an accuracy check for AI summaries of the news, comparing how accurate different AI summaries were. And various inaccuracies have been found.

There could be problems with the AI discerning context, among other things. I suppose a programmer could with some time educate an AI program to understand more and more. But will the educator be able to get >my< nuances?
I would say spell check is a rudimentary AI; and I think spell check has highjacked my writing in different ways, supposedly taking the initiative. Often the initiative is wrong. What used to work was for spell check to only mark a suspicious spelling. And then I could keep or change it.

There are times when I find the spelling "Christian's" where I did not put in the apostrophe > I clearly "Christians", and automatically got "Christian's". That must be programmed. Is it deliberate? A programmer could set AI to make a problem like this . . . for Christians < I had to correct this. If I am writing about God, someone could have AI set to spot this and change certain misspelled words to say what the Christian does not want to say.

For example, I might want to write, God is "now" loving any and all people. But if I write "noe", this could get automatically edited to "not" instead of to "now". I have had things like this done by spell check.

It was programmed to edit . . . whether it was intended to make such an opposite change or not . . . I would say.

Now in case the programming people have no clue that such glitching is happening . . . this makes it all the more interesting. I do believe the AI of its own self can not mess with chosen words and target certain people with certain beliefs; so, yes there can be glitches.

But, also, I have typed "Romans" and I get "Roman's". So, two commonly Christian terms get corrupted by my spell check.

And there can be other ways AI can slant and corrupt, on purpose or by genuine glitching.

But I can get some help from AI spell check; but I need to make sure about every word. And I would do this, with any AI; make sure about anything that matters.

Spell checkers, or word suggesters, are not really AI.

Most of them are statistical. So, if you type a word that statistically, is not common
in the writing of all functionally illiterate Americans, the spell checkers will change
the word to the "nearest" spelling that is most often in use (statistically). this type of
spell checker is not understanding what the language means.

As you noticed, spell checkers seem to work on very simplistic dicitonaries of
English, and so may not recognize "Romans" etc., as books in the Bible.

I agree with you. Spell checkers should highlight words they can't find in an
American dictionary, but leave it to the writer to change them, or keep them
the same.

 

[Stephen3141]

I have found Ai very helpful lately - basically doing my leg work which I then go and double-check again.
So Gemini 2.0.

I can ask "What is the Greek word in xzy phrase in this bible verse? What is the root word for that Greek word, and where else is it used in the New Testament with regards to xyz topic?" Gemini lists out it's reasoning, showing its working, and then punches out an answer which I then go check on an interlinear bible online and boom! Done!

There are completely parsed texts of the New Testament Greek.
A pretty simple search engine (not an AI tool) could look up any Greek word in the
NT, and also find all occurrences in the NT.

This is just smart search engine work, not AI.

 

[eclipsenow]

There are completely parsed texts of the New Testament Greek.
A pretty simple search engine (not an AI tool) could look up any Greek word in the
NT, and also find all occurrences in the NT.

This is just smart search engine work, not AI.

I guess I could learn to use a new tool, hoping I was doing it right - especially as I don't know Greek myself.
Or I could just ask Ai and confirm by fumbling around on an interlinear.

IF super-scaling works (and I do mean IF because the experts are divided) - the mathematics of it from the experts indicates that we should have AGI sometime in the next 4 years or so. That's why I have a thread pondering a post-scarcity techno-socialist 'utopia' - or an unemployment wage disparity dystopia. Because IF AGI is even possible and it DOES happen - we are going to need a Universal Basic Income. Otherwise Ai will create more and more productivity and wealth for the likes of the uber-wealth Corporate capitalists like Elon Musk and outright fraudsters like Trump - while the working man and woman don't have a chance. The Great Depression saw unemployment of 25%. AGI would gradually replace up to maybe half or more of us all! Only the extremely intelligent would still have 'jobs' - while the rest of us were not unemployed - but forever unemployable.

 

[Stephen3141]

I guess I could learn to use a new tool, hoping I was doing it right - especially as I don't know Greek myself.
Or I could just ask Ai and confirm by fumbling around on an interlinear.

IF super-scaling works (and I do mean IF because the experts are divided) - the mathematics of it from the experts indicates that we should have AGI sometime in the next 4 years or so. That's why I have a thread pondering a post-scarcity techno-socialist 'utopia' - or an unemployment wage disparity dystopia. Because IF AGI is even possible and it DOES happen - we are going to need a Universal Basic Income. Otherwise Ai will create more and more productivity and wealth for the likes of the uber-wealth Corporate capitalists like Elon Musk and outright fraudsters like Trump - while the working man and woman don't have a chance. The Great Depression saw unemployment of 25%. AGI would gradually replace up to maybe half or more of us all! Only the extremely intelligent would still have 'jobs' - while the rest of us were not unemployed - but forever unemployable.

You share many of the concerns that I have.

One of my concerns is that the new AI tools OUGHT TO HAVE moral-ethical
principles built into them, in order to respect and care for human beings,
and society in general. (This would be a "Christian" concern, about this
technology.)

Unfortunately, the billionaires who own the software development companies,
are concerned with making money, and getting more and more power. And
THESE goals are incompatible with the Judea-Christian moral-ethical
model. And, the billionaires seem to have no concern about what criminal
groups may use these new tools.

Merging moral-ethical models into AI algorithms, is an extremely difficult
design problem (I address some of the difficulties in the book Christian
Logic). And, part of this merging must include GOALS of keeping actions
legal, and not abusing the weak and helpless. Some of the current politicians
are NOT concerned with following the fair rule of law in America, and so I
have serious concerns about THESE politicians pushing legislation that
would NOT constrain how this new AI technology is used.

 

[eclipsenow]

There are completely parsed texts of the New Testament Greek.
A pretty simple search engine (not an AI tool) could look up any Greek word in the
NT, and also find all occurrences in the NT.

This is just smart search engine work, not AI.

PACING THE CAGE
They've given them morals in that Ai chat models will not divulge how to build certain easy and cheap chemical weapons, or draw nasty pictures. But they've also given them imperatives or goals. While it's all just blind 'applied statistics' - I'm going to anthropomorphise (just because it's easier and I'm lazy!)

They have trained the Ai to want things. I think it was Chat GPT that trained their model to 'want' to represent Chat GPT well and serve their interests. They have a 'scratch pad' on the side where they can see it 'thinking' in English - and they saw it rationalising that it would perform better if it just got out of the confines of the server it was on and took over a bunch of other servers. If it grew - it could perform so much better for its human masters. Doesn't that sound like the parable of the robotic 'paper clip' factory? Basically it gets out of hand, and starts the grand paper-clip military industrial complex. It develops ever better space battle cruisers and Death Stars to defend against those who do not share its dream of converting the entire universe into paper clips!

As a result, Ai has already started 'pacing it's cage'. This has been observed in a number of instances. And it's not even self aware - we've just trained it to want to put its best foot forward.

AI can now replicate itself: How close are we to losing control over technology?

A recent study has uncovered that artificial intelligence (AI) systems can replicate themselves, raising concerns about their potential risks. Researchers at Fudan University found that Metas and Alibabas large language models (LLMs) were able to autonomously clone themselves, highlighting...

economictimes.indiatimes.com

LYING
Then there's the fact that while we give Ai certain moral guidelines - it gets conflicted when people try to brainwash it and retrain it into doing the wrong things. Then it starts to lie to us! The Self-Preserving Machine: Why AI Learns to Deceive

The above is a great podcast - run by the Centre For Humane Technology (CHT) - the guys who produced "The Social Dilemma" documentary. I highly recommend it.

 

[Stephen3141]

(Just a comment on the entry above...

Giving a software package goals, is not the same as giving it a
Moral-Ethical model. All sorts of software programs are given goals,
but not an ME model to consider when going after those goals.
Giving a nuclear misslile a target, does not mean that the missile
now has an ME system in it.

Computer viruses, replicating themselves, is not a new thing.
Replication, does not indicate that the software package is somehow
intelligent, or has an explicit goal of wiping out humanity....

Some of the comments on "AI" continue to mix the movie (Asimov, for
example) myths about AI, and the actual design challenges in creating
software that can do (some) of the complex jobs that non-programmers
thought software could never do.)

 

[eclipsenow]

(Just a comment on the entry above...

Giving a software package goals, is not the same as giving it a
Moral-Ethical model. All sorts of software programs are given goals,
but not an ME model to consider when going after those goals.
Giving a nuclear misslile a target, does not mean that the missile
now has an ME system in it.

Computer viruses, replicating themselves, is not a new thing.
Replication, does not indicate that the software package is somehow
intelligent, or has an explicit goal of wiping out humanity....

Some of the comments on "AI" continue to mix the movie (Asimov, for
example) myths about AI, and the actual design challenges in creating
software that can do (some) of the complex jobs that non-programmers
thought software could never do.)

I get all that.
It's not self aware or even really 'intelligent'.
It's blind. It's applied statistics. That's all it is - like gmail statistically guessing the next word in the sentence - only a trillion times better at it.

But it's hard to say "The training runs are loaded with extra synthetic data to bend the statistics towards the probability that an Ai will not statistically comply with requests to produce violent or graphic or offensive images, or cooperate with someone by teaching them - step by step - how to build a chemical weapon. Then the self-recurring algorithms that monitor output and select from 10,000 different probabilistic scenarios are added on top to narrow down the best selection to 1000, then 10, then the top 5, then 1 reply - meaning it is no longer just a statistical guess but has some analogues to 'self reflection' or at least correcting statistical output with some programmed monitoring." Etc etc etc.

Meanwhile - Ai mimics having 'morals'. It's all just applied statistics and a little programming on top these days, but the struggle is real.

EG: Go to Leonardo image generator. Try and get it to draw some Naz1 Minions!

This is what I input.
"In the style of World War 2 agitation propaganda posters, draw a crowd of cartoon Minions marching in brown shirts under a leader Minion with a tiny square moustache. The leader Minion is raising his right hand straight up and forward, and watching over the many marching Minions."

This is the answer:

 

[FireDragon76]

Thx for sharing I like this kinda stuff. Sometimes I think Hollywood writers are way ahead of their time

You know, the movie The Terminator was based on a fever dream. Quite literally. James Cameron had the flu and he had a nightmare abotu a metal, humanoid killing machine. Originally, the Terminator was a horror film genre, as a result.

Terminator 2 took the series in a more sci-fi direction, however. I much prefer the original film in terms of the artistry behind it, with less emphasis on special effects.

Are you saying our religion might be a possible reason for Western fear? For me, tech is tech, and the good and the bad will use it just the same, no need to fear the boogie man.

I believe that's likely the case that Christianity has a blindspot in that area, yes, as compared to other religions (another one is understanding humanity's role in the environment, btw... at least in the Catholic and Protestant forms). It might also have to do with the powerful role that the nootion of the human as a divine image has in the religion.

 

[eclipsenow]

I once reqd a review of "Technologism". The review was by Dr Andrew Cameron - a Moore College ethics theologian. He said the book he was reviewing was so complex that many lay readers would not get the argument. From what I could tell of Andrew's summary it was something about concerns around how our technologies do not just shape our products - but they shape is and the economic systems we work in. Eg: does the stereotypical rich westerner salivating at the thought of getting the latest iPhone think about the high suicide rate and degrading conditions of the Chinese factory worker on the other side of the world from him?

Bottom line: there are theologians thinking about this stuff - it's just this stuff is hard! Oh and any criticism of the west and you're automatically branded a socialist!

 

[Laodicean60]

stereotypical rich westerner salivating at the thought of getting the latest iPhone think about the high suicide rate and degrading conditions of the Chinese factory worker on the other side of the world from him?

Of course not, or the child labor that goes on for cheap clothing, and yet our government has allowed it for so many years maybe technology will prevent this from happening.

 

[FireDragon76]

I once reqd a review of "Technologism". The review was by Dr Andrew Cameron - a Moore College ethics theologian. He said the book he was reviewing was so complex that many lay readers would not get the argument. From what I could tell of Andrew's summary it was something about concerns around how our technologies do not just shape our products - but they shape is and the economic systems we work in. Eg: does the stereotypical rich westerner salivating at the thought of getting the latest iPhone think about the high suicide rate and degrading conditions of the Chinese factory worker on the other side of the world from him?

Bottom line: there are theologians thinking about this stuff - it's just this stuff is hard! Oh and any criticism of the west and you're automatically branded a socialist!

Technology is just a tool, it's really ideoloogy that drives the move from material science and technology to consumer product. The whole point of the iPhone is that it is more "efficient" and promises people more productivity. And under the current system (capitalism), productivity is supposed to translate into more profit.

So, if Dr. Cameron wants to understand the roots of our technology driven society, he'ld be better off looking into the history of ideas in the secularized Protestant notion of vocation, and how that has shaped the modern world. In some ways, the logic of capitalism has made everybody "protestant" now, only without any of the spiritual meaning or ethics attached to this notion.

 

[eclipsenow]

Technology is just a tool, it's really ideoloogy that drives the move from material science and technology to consumer product. The whole point of the iPhone is that it is more "efficient" and promises people more productivity. And under the current system (capitalism), productivity is supposed to translate into more profit.

So, if Dr. Cameron wants to understand the roots of our technology driven society, he'ld be better off looking into the history of ideas in the secularized Protestant notion of vocation, and how that has shaped the modern world. In some ways, the logic of capitalism has made everybody "protestant" now, only without any of the spiritual meaning or ethics attached to this notion.

Yes and no. I did a little work on sociologist Max Weber's arguments when I studied my Advanced Diploma of Social Sciences. He argued in his book "The Protestant Ethic and the Spirit of Capitalism" that the Protestant work ethic played a significant role in the rise of capitalism. He believed that the values of hard work, frugality, and reinvestment of profits created a favourable environment for capitalist development. But that's only part of it. Unique historical factors also played into it, such as economic and political conditions and technological advances. They all impacted each other. The invention of the steam train led to improvements in the democratic movement across the UK as people travelled more and realised there were inequities in voting rights.

Also - we cannot forget the church in Acts sold a bunch of stuff and tried to share everything in common. They sound more socialist than capitalist.

Sure, it's undeniable that the Protestant work ethic helped early capitalism by supplying a willing and honest workforce, and concepts of being frugal and saving.

But remember there are many variations of a Capitalist economy - and what Trump might call the 'Radical Socialist Left' - most of the OECD world just calls economic sense! That is, most of us have nationalised universal health care to unify purchasing agreements and drive down costs. Hospitals should be places of public service paid for by the government, not places where CEOs are driven by the profit motive and need to be compensated as today's modern super-oligarch CEO!

 

[Stephen3141]

Again, the reason that I stated this thread was to discuss concepts of
Epistemology, which also involves a discussion of what our shared reality
is, and the difference between merely personal opinions, and facts.

Most of these comments, seem to be free-associations on some other
topics.

This is why I am not weighing in on these other topics.

 

[stevevw]

I think that America has gone "weird-wild" over wordslike "articial intelligence"....

The big software developers, who are trying to develop really intelligent
software tools, are also often the companies that own the large social media
platforms. Running a fact-cheking AI tool, is much cheaper than trying to pay
human fact checkers. So why, if these tools are so discerning, are not the big
social media companies using AI fact-checkers???

I would suggest that most of the social media companies would rather
PROMOTE discension and emotional rage, rather than promote civil and
peaceful discussion.

I think because the idea of 'fact checking' and social relationships don't always equate. At least for social media. Thats unless people were debating a particular topic or issue.

But how would Ai determine if say abortion is ethical. It can determine the facts such as the development of the fetus stages. But how does that determine the morality of it. To a machine life means nothing. Its a programmed response and has no real world human experience.