The Monster at the End of the Book

by Dick Hall-Sizemore

We have created a monster. The genie is out of the bottle. Whatever metaphor you want to use, there is no going back and the way forward poses great dangers. The monster or genie is AI.

The media are full of the promising possibilities of AI improving our lives—great leaps in medicine, science, technology, manufacturing, etc. There is less discussion of the effects of having leaders in business and government, as well as the bureaucrats in those spheres, who are incapable of composing a coherent paragraph on their own due to their reliance on ChatGPT through high school and college.

What really concerns me is the potential of AI for politics; elections in particular. Candidates, or, more likely, sympathetic groups, could release recordings having opposing candidates seeming to say what they did not say. For example, residents of New Hampshire recently received a robocall with what sounded like the voice of Joe Biden urging them to skip the primary election there.

For the upcoming election, think of the effect of a video surfacing that showed Trump doing what the Steele dossier alleged he did in Russia. There is no doubt this is possible. After all, last fall, a group of teenage boys in New Jersey, being teenage boys, circulated pictures of girls in their classes with nude bodies (not theirs). Experts say that all it takes is an iPhone and easily accessible AI software.

I have been mulling all this over for a while. It turns out that I was not thinking broadly enough. In addition to being a possible weapon, the existence of AI provides politicians “plausible deniability,” as one expert explained. FOX News recently ran an ad comprised of well-documented gaffes of Trump. He responded, “The perverts and losers at the failed and once disbanded Lincoln Project, and others, are using A.I. (Artificial Intelligence) in their Fake television commercials in order to make me look as bad and pathetic as Crooked Joe Biden, not an easy thing to do.”

I should not have been surprised at this response. After all, this is a man who insists, in the face of mountains of evidence to the contrary, that the 2020 election was stolen. And millions of people accept his version. I expect that, at some point in the upcoming year, Trump or some group allied with him, will be claiming that all the footage showing the January 6 attack on the Capitol was AI-generated.

It used to be said, “Seeing is believing.” That is no longer true. We are entering a world in which we will not know what to believe. We will not know whether to believe that what we see, pictures, video, film, etc., is real or AI-generated. Truth will become elusive. Or, perhaps truth and reality will cease to exist as objective concepts and become whatever one defines it to be at that moment.


Share this article



ADVERTISEMENT

(comments below)



ADVERTISEMENT

(comments below)


Comments

98 responses to “The Monster at the End of the Book”

  1. Chip Gibson Avatar
    Chip Gibson

    Though very much at odds politically, I very much share your concerns on this topic. Nice article, Sir – would hope that we overestimate the danger – but, hope is not the answer to this real threat. Your article should be taught in every grade school.

  2. LarrytheG Avatar

    The thing is, as you noted, we can’t go back. We’re not going back.

    We have to learn how to deal with it and it’s a pretty tall order
    given our current inability/refusal of some to deal with truth and realities, and believe in and promote lies and conspiracy theories.

    It’s a powerful tool that can be used for good and/or evil .. and
    will be.

    Good article, thanks for posting. There will be a need for more IMO.

  3. walter smith Avatar
    walter smith

    I’m sorry. The election was stolen.
    Ignore all the statistical anomalies. Ignore the irregularities we saw before our eyes. Ignore the few court cases that have gotten to adjudication to rule that Wisconsin and Pennsylvania changes violated the law. Ignore the common sense question of did Joe Biden really get 8 million more votes than Obama got in 2008?

    Now, ask yourself… if a trustee of a trust fund for your benefit was accused of violating his fiduciary duty, how would that trustee react?
    If a crook, he would not open the books and would do everything he could do to avoid opening the books. If honest, he would immediately open the books and show you. Why has that not happened? To ask the question is to answer it.
    Meanwhile, remember, according to the smart people, it is crazy to think cheating occurs or is possible…
    https://www.thegatewaypundit.com/2024/01/gig-is-up-exclusive-local-reporter-describes-election/

      1. walter smith Avatar
        walter smith

        And that has what to do with what? That the brain dead, corrupt, kinda pedo SlowJoe is the President? Yeah, he is. And he is illegitimate. 51 “national security experts” assured us Hunter’s laptop had all the earmarks of Russian disinformation, even though the FBI knew in 2019 it was real, and the data on it matched his iCloud backups which they had obtained months earlier in 2019. I’m not the fool.
        And it is insulting how cheaply the brain dead corrupt head of the Biden crime family sold us out for.
        And yet not one Democrat pol has the integrity to condemn the bribery, lying, lawfare, by any means necessary Will to Power… But what else would you expect from a party whose primary raison d’etre is maximizing the killing of babies?

        1. Nancy Naive Avatar
          Nancy Naive

          Existence of a vulnerability is not proof of an explotation.

          1. walter smith Avatar
            walter smith

            Yes. But remember, we were told absolutely safe! Can’t be hacked! Now you move the goalposts…again…and will continue to do so. Much like correlation is not causation I heard all during Covidiocy. But it could be…and just so many coincidences… How did they happen?

    1. Teddy007 Avatar

      Citing gateway pundit is not way to convince anyone.

      1. walter smith Avatar
        walter smith

        Ah, yes, the I’m not gonna look strategy. You are the duped one already, and will only continue to be duped as you are spoonfed the propaganda you want. The article you won’t read is about a trial in Ga where the Dominion machine was hacked…in court… with nothing in particular. The lawyer used the other lawyer’s pen to do it. But you stay ignorant…on purpose…
        I think the NY T is untrustworthy. RUSSIA RUSSIA RUSSIA? Pee tape? Anyone? Hunter laptop? Bat cave origin 1000 miles from Wuhan?

          1. walter smith Avatar
            walter smith

            So you posted the article to explain why you are always wrong? Or is this another Larryesque posting of a headline without actually reading the drivel posted? You did read this of course? And all of the political “scientists” are truly just data driven “scientists” and not political hacks? How did they determine what was true and what was misinformation?
            Let me give you a test to see if you are spreading misinformation – how many sexes are there?
            I’ll check back in a week or so.

          2. Not Today Avatar

            How many conservatives does it take to nominate someone other than a sex offender? Don’t worry, I’ll wait.

  4. James Wyatt Whitehead Avatar
    James Wyatt Whitehead

    Mr. Dick the 21st century is going to be hard on guys like you and me.
    https://uploads.disquscdn.com/images/818dfc7843f01741afd16c805c16bc0137dbe5f42abe5d7291d283fe5550427c.jpg

    1. Nancy Naive Avatar
      Nancy Naive

      Guys with prison tats?

      1. LarrytheG Avatar

        ouch! yes!

      2. James Wyatt Whitehead Avatar
        James Wyatt Whitehead

        You should know this one sailor! From the movie Master and Commander.
        https://www.youtube.com/watch?v=azMA6upM5-Q

        1. Nancy Naive Avatar
          Nancy Naive

          So, not prison. A fo’castle rat?

          Good movie. Some plagiarism from other stories. But, all-in-all entertaining.

  5. Nancy Naive Avatar
    Nancy Naive

    Look on the bright side. We are leaving the Age of Belief and entering the Age of All Knowledge. We no longer have to believe without evidence; we can produce the evidence for what we know.

    1. Kathleen Smith Avatar
      Kathleen Smith

      Scary

      1. Nancy Naive Avatar
        Nancy Naive

        Maybe. Imagine the boon to law enforcement. VB cops showed suspects faked DNA reports to elicit confessions. Imagine how many confessions they’ll get when they show their suspect security video of him shooting the clerk?
        “Boy, that’s you! No juror will see that and not convict you and that’ll put the needle in your arm, but if you confess, maybe the CA will take the death penalty off the table.”

        Every murder gets solved.

      2. Nancy Naive Avatar
        Nancy Naive

        Maybe. Imagine the boon to law enforcement. VB cops showed suspects faked DNA reports to elicit confessions. Imagine how many confessions they’ll get when they show their suspect security video of him shooting the clerk?
        “Boy, that’s you! No juror will see that and not convict you and that’ll put the needle in your arm, but if you confess, maybe the CA will take the death penalty off the table.”

        Every murder gets solved.

      3. Nancy Naive Avatar
        Nancy Naive

        Maybe. Imagine the boon to law enforcement. VB cops showed suspects faked DNA reports to elicit confessions. Imagine how many confessions they’ll get when they show their suspect security video of him shooting the clerk?
        “Boy, that’s you! No juror will see that and not convict you and that’ll put the needle in your arm, but if you confess, maybe the CA will take the death penalty off the table.”

        Every murder gets solved.

        1. Kathleen Smith Avatar
          Kathleen Smith

          Even scarier!

        2. LarrytheG Avatar

          what about meta data?

          1. The rule of “garbage in, garbage out” will still apply.

          2. LarrytheG Avatar

            you beat me to it… yes…. but also things like what happens when you multiply/divide one big number by
            another that is .001 E-10….

        3. DJRippert Avatar

          Or, ….

          1. Use cold case DNA to create a facial drawing of the person whose DNA it is.

          2. USe facial recognition to match the DNA-based drawing to a real person.

          3. Arrest the real person.

          The company that makes this technology is based in Reston, Va.

          https://www.wired.com/story/parabon-nanolabs-dna-face-models-police-facial-recognition/

          1. Nancy Naive Avatar
            Nancy Naive

            Been around awhile too. Some success.

  6. Nancy Naive Avatar
    Nancy Naive

    Don’t believe anything you read and only half of what you see. — Will Rogers. Adapt or die. We’ve long ago overcome nature. Now we need to overcome nurture.

    As for AI, it’s a tool. I have a pocket calculator on my desk. I use it. I could use pencil and paper, but I don’t. That I can do long division doesn’t mean that my calculations are better than someone who can’t. He has a pocket calculator too.

    I foresee the day when droids bid in cryptocurrency for non-fungible tokens. But there’s an upside. You won’t get 4 Supreme Bots voting to nullify the Supremacy Clause.

    1. That I can do long division doesn’t mean that my calculations are better than someone who can’t. He has a pocket calculator too.

      Perhaps, but knowing how to do long division allows you to more easily detect any data entry errors you may have made. You can look at the answer the calculator gives you and determine whether it makes sense. A person who cannot do long division is more likely to blindly trust the calculator’s results.

      1. LarrytheG Avatar

        re: ” A person who cannot do long division is more likely to blindly trust the calculator’s results.”

        even if they do it twice?

        😉

        1. As long as one recognizes his/her arithmetical limitations (e.g. doing backup computations to assure accuracy) a calculator is a fine tool.

          Personally, my limitation is not math related, it is that I am somewhat “fumble-fingered” when using a keyboard. I usually have a pretty good idea of what my calculator’s answer should be, but I do the computations twice anyway just to make sure I didn’t hit the wrong key at some point during the entry process.

          1. LarrytheG Avatar

            having some idea of what the ballpark answer ought to be…. maybe? Long ago, had a job that required doing “hand cases” to verify a computer-generated algorithm answer. Needed to use Marchant. The hand-case could take all day. There would be two of us doing it so we had to both us and the computer match. That job was called a “Math Aid”. 😉

          2. You have my sympathy. I hope you did not stay at that particular job very long.

          3. LarrytheG Avatar

            It was a test. If you could do that, then a path was open for advancement. If you could not
            do it, then your “job” at that level was assured…. trying to match a computer model of a point
            object moving a 3d axis was a challenge but it taught huge respect for the power of computers
            in general but also… garbage in , garabage out.

  7. LarrytheG Avatar

    AI can be somewhat compared to drones. Modern drones have totally changed the world in many aspects – for good and bad.

    But I still point out that even before the advent of AI, there was a problem with dealing with truth and realities not only elections, but pandemics and climate , pizza shops, moon landings, Holocaust, slavery, history in general, etc.

  8. Teddy007 Avatar

    Anyone who thinks that the good old days were better are fools. Politicians used to say one thing in a speech to one group and something different in a different speech. The in person reporters knew what was going on but did not care.
    Reporters did not care how bad and corrupt politicians were in the past. Reporters did not care about how immoral politicains were in the past. Can anyone imagine someone like Wilbur Mills existing in an age of social media and ever present camera?

    1. Dick Hall-Sizemore Avatar
      Dick Hall-Sizemore

      There were some reporters in the early 1970’s who cared about corrupt politicians, as Richard Nixon to his sorrow found out. Yes, I can imagine someone like Wilbur Mills existing today. He is running for President,

  9. DJRippert Avatar

    AI is pretty much all I do these days. Deepfakes are an issue. They are getting more realistic. Here’s an example that was created 2 years ago:

    https://youtu.be/oxXpB9pSETo?si=Nh2yWO_vFCfOAdzn

    But, notice there is no background behind fake Morgan Freeman and the video doesn’t show his body – especially his hands. Image generators have considerable issues (as of today) with accurate portrayal of hands, for example. There are often 6 fingers.

    So, if a headshot video were released – even if it was as good (or better) than the Morgan Freeman deepfake, and it went viral – people would immediately ask where it was shot, when was it shot, etc.

    If I were planning to disrupt an election (and I’m not), I wouldn’t want my disruption to “go viral”. Anything that goes viral gets too much scrutiny.

    If I were trying to disrupt an election (in 2024), I’d make 4 bets:

    1. The number of people watching streaming services that are wholly or partially supported by ads is growing by leaps and bounds.
    2. Since you have to login to a streaming service, the streamer knows a great deal about you. They provide as much information as legally possible to advertisers since they can sell ad impressions for much more money is the ads are personalized.
    3. The advertiser (in this case a political campaign) matches the data provided by the streamer to date they have collected over the years. While they may not know the viewer’s name, they know enough to microtarget the viewer.
    4. AI allows the creation of content that can be used in dynamic advertisements. These ads use pre-fabricated components which are constructed immediately prior to display. Which components to use are determined by the data described in step 3 above.

    The net result is that a candidate might display an ad that emphasizes support for farm subsidies in rural Wisconsin while emphasizing his or her opposition to deficit spending in suburban Milwaukee.

    Given that only 7% of adults have a ‘great deal’ of trust in news media, (a recent Gallup poll found) while 38% say they have none at all, those commercials might form the majority of the opinion a voter has about a candidate.

    You pollute the minds of voters the same way Tik-Tok pollutes the minds of kids – by feeding them a continuous stream of what they want to hear, whether it truly represents the candidate or not.

    1. LarrytheG Avatar

      I was with you until the end part about trust in the media and you ignored the FOX News/Daily Wire/etc aspect.

      1. DJRippert Avatar

        They’re not trusted either. People gravitate to the outlets that tell them what they want to hear because they don’t trust media (from either side) to provide an unbiased view. So, they see ads that tell them a fraction of what a candidate says (even if it contradicts other things said by that candidate). They do not verify the claims with the media because they do not trust the media.

        https://www.poynter.org/commentary/2023/american-trust-in-media-is-near-a-record-low-study-finds/

        1. LarrytheG Avatar

          Who is Poynter and why should I trust them? And why do you trust them?

          1. DJRippert Avatar

            Don’t trust Poynter, trust the Gallup Poll referenced in the Poynter article.

          2. LarrytheG Avatar

            so.. why trust GALLUP? Many of the folks who don’t trust media also don’t trust GALLUP or PEW, right?

          3. DJRippert Avatar

            You are exactly the kind of voter I would aim those ads at.

          4. LarrytheG Avatar

            but I explicitly don’t believe conspiracy theories or FOX news or the DailyWire, etc… so
            why do I, should I , “trust” GALLUP and PEW? There are a ton of polling companies that
            do really dishonest polls, right? So why trust these two over those others?

          5. DJRippert Avatar

            I try to read both conservative and progressive news outlets and make my own decision as to which is more accurate.

          6. LarrytheG Avatar

            How about the folks that the banks and Credit Unions have to warn about talking to scammers and
            giving their personal data and taking money out to get to the scammers? How do those folks do
            with “news” and “media”?

          7. DJRippert Avatar

            Not well, which is why I’d try to rig an election quietly rather than with a deepfake October surprise.

          8. LarrytheG Avatar

            I figure the ones that have to be told how to protect their data and bank account are probably just as easily influenced on other issues including politics and elections.

            Given Mr. Trump’s dishonest and misleading claims and the fact that his supporters apparentl believe them…
            probably a high number also susceptible to other scams.

      2. He didn’t mention any specific media outlets by name. He was completely non-partisan in his comment.

        So, why did you feel it necessary to [once again] make known your personal partisan hatred of Fox News and Daily Wire, etc.?

        1. LarrytheG Avatar

          I don’t have a hatred of them. I just think they lie out the wazoo…

          but his reference:

          ” This poll is the first time the percentage of Americans with no trust in the media at all is higher than the percentage of respondents with a great deal or fair amount of trust combined.

          The partisan divide remains strong with media trust, with 70% of Democrats saying they have a great deal or fair amount of confidence in the media, while 14% of Republicans and 27% of independents say the same.

          A majority of Republicans indicated they have no trust at all in the media for the third straight year.”

          but those same Republicans apparently do trust FOX and other right wing “media”.

          Right?

          So maybe some confusion here over the word “media” and whether or not FOX and allied are also “media”?

    2. LarrytheG Avatar

      re: ” Since you have to login to a streaming service, the streamer knows a great deal about you. They provide as much information as legally possible to advertisers since they can sell ad impressions for much more money is the ads are personalized.”

      Didn’t you also just describe what TikTok essentially does?

      1. DJRippert Avatar

        Yes. And Facebook, etc.

  10. DJRippert Avatar

    ChatGPT is essentially an auto-complete application (like on your phone when you are writing a text) on steroids. Those steroids are:

    1. Immense amounts of training data.
    2. Almost inconceivable amounts of compute power.
    3. A very clever mathematical approach called the transformer architecture (first described in a Google white paper, 2017).

    ChatGPT doesn’t think. It doesn’t reason. It generates text by probabilistically “guessing” the next word based on (among other things) the common crawl, which is a scraping of the entire public internet.

    ChatGPT4 passed the New York Bar exam. It took an IQ test and rated out at the same level as Elon Musk. It got over 1400 on the SAT. Oddly, it failed all five parts of the CPA exam.

    If an auto-complete application on steroids can excel at standardized tests, I have to wonder about the value of those tests.

    The prior blog post was titled, “Failure is not an Option with Proposed SOL Revisions”.

    I wonder how ChatGPT would do with the SOLs.

    The better it does, the less value I see in those SOLs.

    1. LarrytheG Avatar

      What is the value of asking ChatGPT for the references it consulted in generating it’s response?

      1. DJRippert Avatar

        Let’s ask ChatGPT …

        “As an AI developed by OpenAI, I generate responses based on a mixture of licensed data, data created by human trainers, and publicly available data. I don’t have the capability to access or cite external sources or databases directly when generating responses. Instead, my responses are based on the knowledge I’ve been trained on up to my last update, which includes a wide range of texts and materials.”

        The references you sometimes see cited are from something called Retrieval Augmented Generation (RAG). That’s what you get when you ask an LLM like ChatGPT or Google’s Bard to use a search engine to augment its response. Bard uses the Google search engine, ChatGPT usually uses Bing.

        ChatGPT’s last training update was April, 2023 as I recall. Anything added to the internet since then is not in ChatGPT’s LLM. So, adding search brings updated data into the conversation. I believe it is those search inquiries that are cited as references.

        1. LarrytheG Avatar

          Provided references are a way to confirm/validate/assess the information provided both manually as in Wiki and Peer-review/related and in automated bots that “search” for their references.

          You can ask the bot about their references. You can question them and ask for more or provide one
          it did not.

          You can decide if the BOT is credible or not by asking questions of it which is way better than some non-bot sources that make claims!

          1. DJRippert Avatar

            ChatGPT doesn’t know what sources were used to train the specific weights in its neural network any more than you or I remember who taught us that Washington, DC is the capital of the United States. Or, how we know that knife starts with a “k”.

          2. LarrytheG Avatar

            You’re talking about how ChatGPT was actually programmed…correct?

            If you ask ChatGPT the same question you asked Google Bard or one of the other AI bots…would
            that be useful for validating their performance no matter how each might have been programmed?

          3. DJRippert Avatar

            As an AI – ChatGPT is more trained than programmed.

          4. LarrytheG Avatar

            the core of it is software, right? “ChatGPT is a natural language processing tool driven by AI technology that allows you to have human-like conversations and much more with the chatbot. The language model can answer questions and assist you with tasks, such as composing emails, essays, and code.”

            Have you heard of a LALR parser? https://en.wikipedia.org/wiki/LALR_parser

  11. Ronnie Chappell Avatar
    Ronnie Chappell

    Sorry, but we reached the point where people believe very little of what they read, see and hear in the media. Not sure AI can make it much worse.

  12. Or, perhaps truth and reality will cease to exist as objective concepts and become whatever one defines it to be at that moment.

    We don’t need AI for that. We’ve been heading in that direction at a pretty good clip for quite some time. It was definitely underway during the days of “it depends on what the meaning of is, is” and even before then.

    The term “situational ethics” also comes to mind. And then there’s the conspiracy theories surrounding the moon landing, MLK’s murder, the Kennedy assassinations, etc., etc.

    1. Lefty665 Avatar

      There was a cartoon (Washingtoon – Mark Samaty) in the Reagan years that featured Reagan’s perceptionometers. They were people whose job it was to generate favorable perceptions. Their motto was “Issues are not real, they are just masses of opposing scenarios vying to become perceptions.”

      One conspiracy theory that caught me was at the time of the first moon landing, a woman at a gas station told me that she was against that. She was afraid the rockets were punching holes in the atmosphere and that all our air would leak out. I suggested she be sure to fill her tank so she would be prepared to drive to where there was more air and to roll her windows up so what air she had would not leak out.

      What good is a conspiracy theory if you can’t feed the paranoia? In that case trying to explain gravity was clearly going to be a non starter. It also may be an indicator of how horrible Richmond’s schools have been for a long long time.

  13. LarrytheG Avatar

    I predict the development of “Truth Bots” as well as a law that requires any AI-generated information to be identified as to it’s origin or it will be treated as questionable and not authentic.

    1. Dick Hall-Sizemore Avatar
      Dick Hall-Sizemore

      How will we know if something is AI-generated or not?

      1. Nancy Naive Avatar
        Nancy Naive

        The Bot will. Takes one to know one.

        1. LarrytheG Avatar

          Yep. There will be “truth” bots that do things like checking the meta data – that will be required by law an if missing, it is not an authentic source.

          It will be a more comprehensive process than, for instance, checking the URL of a source to see if it came from a credible source.

          If you query BARD or CHATGPT on a subject, it will provide sources and references and you can ask that they do also.

          1. Nancy Naive Avatar
            Nancy Naive

            Of course, SCOTUS will determine such bots to be an invasion of privacy.

          2. LarrytheG Avatar

            SCOTUS is proving itself unable to deal with the realities of the 21st century just like
            some elected lawmakers who don’t understand that TiKToK is just a name of a particular
            kind of software that can be replicated and named something else. If you tell them this,
            they go into blue-screen mode.

          3. Nancy Naive Avatar
            Nancy Naive

            Once they overturn Chevron, then they will be the sole source of all science, the court of clerics so to speak, and it won’t matter.

          4. LarrytheG Avatar

            talk about DUMB! I want to see them weigh in on the IRS Tax code and regulations!

          5. Nancy Naive Avatar
            Nancy Naive

            Why would they care about the IRS codes? You don’t declare gifts.

          6. LarrytheG Avatar

            oh the idea that the IRS sets rules and regulations based on the law?

            I want to see SCOTUS declare that to be illegal and every single thing in the IRS code has to be written in the law.

            That’s on top of the idea that the justices themselves are apparently above the law on receiving undisclosed gifts and related. And no law regulates that for sure!

          7. Lefty665 Avatar

            Can’t be any worse than “science” by Fauci and CDC (what a profound disappointment – they used to be the gold standard)) and might be better.

            An old fella I once knew used to would say “Can’t hurt none and it might help some” when asked about doing something to fix a problem. But he was from Carolina, so go figure.

          8. Lefty665 Avatar

            Can’t be any worse than “science” by Fauci and CDC (what a profound disappointment – they used to be the gold standard)) and might be better.

            An old fella I once knew used to would say “Can’t hurt none and it might help some” when asked about doing something to fix a problem. But he was from Carolina, so go figure.

          9. Nancy Naive Avatar
            Nancy Naive

            Yeah, it will be.

  14. Nancy Naive Avatar
    Nancy Naive

    It’s like the 3D printed gun. People got all excited over nothing. Once the only source of cartridges was HP, no one could afford to shoot one.

    1. DJRippert Avatar

      And my 3D printed gun was perpetually out of cyan bullets.

      1. LarrytheG Avatar

        Given your extensive background in computer technology and on the subject of ChatGPT in general, would you consider writing a blog post on it?

        1. DJRippert Avatar

          I’d like to write such a post but I’m working full time and the post would have to be tailored to Virginia.

          Maybe a post about how Large Language Models could be used in Virginia to make K-12 education more effective.

          If the weather sucks this weekend (and I can’t play golf) maybe I will write such an article.

          Meanwhile, this lecture (about 1 hour) from Mike Woolridge is a good intro …

          https://www.youtube.com/watch?v=b76gsOSkHB4&t=1365s

    2. 3D printed guns were/are definitely “much ado about nothing”, but I’m not sure HP had anything to do with it.

      😉

      1. Nancy Naive Avatar
        Nancy Naive

        Cartridges — another name for bullet. Clearly, you’ve never bought an HP ink cartridge.

        1. I apologize. I should first have complimented you on your clever pun before making my more serious point about the histrionics from anti-gun folks about 3-d printed guns.

          PS – I have bought hundreds of HP ink cartridges – for both printers and plotters.

          1. Nancy Naive Avatar
            Nancy Naive

            I remember buying HP pens for the flatbed and drum plotters. Yeow!

          2. Nancy Naive Avatar
            Nancy Naive

            I remember buying HP pens for the flatbed and drum plotters. Yeow!

          3. Lefty665 Avatar

            Their extended life laser toner cartridges weren’t a bad deal. But then carbon black and a little glue were low cost ingredients.

  15. Lefty665 Avatar

    Worrying about being taken for a ride by AI is fundamentally misunderstanding and under appreciating the threat. It trivializes the issue. The issue is the exponential growth of AI capabilities and the point at which AI decides us meatheads are more trouble than we are worth… Be afraid, be very afraid.

    1. Dick Hall-Sizemore Avatar
      Dick Hall-Sizemore

      70 years ago, Isaac Asimov anticipated this eventuality. His solution was to program every robot, or AI, with the following three “laws”:
      1. A robot may not injure a human being, or, through inaction, allow a human being to come to harm.
      2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
      3. A robot must protect its own existence as long as suh protection does not conlict with the First or Second Law.

      His “robot” novels and short stories are full of instances of how these “laws” conlict and how those conflicts are worked out logically. There is also the rogue robot.

      1. Re Asimov’s robotics Laws, why would an autocratic leader like Putin ever embed a humanitarian-based set of limitations in the AI army he sets about building?

        Another problem: why do you assume that the commentators here calling themselves “Walter Smith” or “Nancy Naive,” to pick two at random, are not long-running constructs of AI?

      2. Lefty665 Avatar

        I grew up reading Asimov, and like his take on things. However, his solution to robotics is, while insightful for the time, now antiquated and naive. Sort of like a Model A compared to a current F1 car.

        If you’re not terrified of AI today you’re either dreaming or not paying attention. Hard to detect fakes are minor symptoms.

        Consider the near future where AI may ponder whether to switch megawatts to a data center to complete its next iteration of self generated super sentience or gate them to the homes of us meatheaded dullards to keep us comfortable.

        Over the years us programmers figured there were 3 levels of programs:
        1) Given good input the program gives the right answer. (aka garbage in, garbage out)
        2) Regardless of input the program will not give the wrong answer.
        3) The program gives the right answer no matter what is input.

        1 is early programming, and 2 is common norms. 3 is rarely achieved.

        AI is rapidly moving beyond this quaint concept. It will be, if it is not already, making decisions with no interest in or consideration of human input.

        Thinking of AI as a tool that humans will use to confuse or influence each other profoundly misunderestimates the genie we have let out of the bottle.

      3. LarrytheG Avatar

        I predict there will be such a thing as trusted Bots and not trustable bots.

        1. And how will you know which one is to be trusted and which one is the Alternative Fact?

          1. LarrytheG Avatar

            you’ll have to decide yourself based on your own experience with the bot
            which is interactive. Sorta like if you can ask several questions of someone
            on a subject, can you decide if they are telling the truth? If a large number
            of people decide they can “trust” or a large number decides they cannot.
            sorta like, do you trust GALLUP or PEW or references in Wikipedia , or OBM
            for budget info, etc… yes.. I realize some folks don’t trust NOAA for climate, etc!

          2. ‘Alternative Facts’ from bots will be the ones which do not comport with his personal opinions, just like now.

          3. LarrytheG Avatar

            Each one will develop a reputation. And you can put the same question to all the bots and see if the answers match or differ.

            For instance, you could ask them all about January 6 and see how they answer and then decide based on their answers.

    2. . The issue is the exponential growth of AI capabilities and the point at which AI decides us meatheads are more trouble than we are worth…

      Like “Skynet” in the Terminator movies…

Leave a Reply