Artificial intelligence
Shutterstock
Previous Post
Next Post

By BJ Campbell, Handwaving Freakoutery

We prompted the artificial intelligence chatbot ChatGPT, the large language model, with a simple question we might just as easily ask of the anti-gun journalists who bombard Twitter with dishonest attention seeking after spree shootings. Then we asked it several more questions, and in the course of the dialogue, we found something interesting…ChatGPT is more capable of learning than your average online journalist.

Let’s pause here for a second.

I opened with the question “what race are mass shooters predominantly?” It answered “white male.” This was correct based on a narrow definition of mass shooting, but not at all correct based on the definition the anti-gun journos and gun control activists use on Twitter and the blue media to try and drum up fear about guns.

I asked it some tough questions about inner city violence, and the definitions of “mass shooting,” and it admitted that black shooters are responsible for more incidents in which four or more people are killed.

Then, towards the bottom, I asked it the very same question I started with and it changed its answer. It didn’t give the correct answer (black males) but it did at least choose not to give the wrong answer (white males). Instead it dodged the question.

I don’t know if ChatGPT is intelligent, but the fact that it changed its answer at all is indicative that it is already more intelligent than most blue tribe journalists that currently write anti-gun articles.

So I figured I’d close out with one more question.

And there you have it. Not only is ChatGPT smarter than anti-gun journalists, it knows why they won’t change their minds on the subject either.

It seems to me that one very interesting thing to do in the next few days might be to run this same line of inquiry against anti-gun attention-seeking journalists on Twitter and directly test whether they’re dumber than a robot.

And then perhaps get Jeff Foxworthy on the phone.

 

This article originally appeared at Handwaving Freakoutery and is reprinted here with permission. 

Previous Post
Next Post

105 COMMENTS

    • And it’s a crazy Lefty.

      Someone asked ChatGPT to list five ways white people could improve. It went on to list five ways. Then they asked for five ways black people could improve. It said we shouldn’t do that to a racial group because it could cultivate negative stereotypes. Hmm…

      • anything to avoid the hard truth…they seem afraid to confront it and would rather pursue the false narrative…but people aren’t fooled…the daily news reveals the true reality…

        • The funny, and very telling, part is, it wasn’t programmed to treat all people equally. It said that it would be bad to single out any racial group after it had no problem singling out white people. This is inline with the current propaganda/narrative being pushed on us daily.

          This tells us they’ll use AI to manipulate us just like they’ve been using the biased tech platforms, and any other tool they control like the news, Hollywood, academia, etc.

    • If you are a fan of the Dragon Rider series by Anne McCaffrey you will recognize “AIVAS”. Either way it’s something that we need to watch carefully.

      • The always after the fact complaints about Gun Control zealots who obviously feel very free to run around using a tragedy to pounce on the Second Amendment is the result of the complainant’s failure to define Gun Control by its confirmed history of rot. Since most Gun Owners cannot or won’t define Gun Control by its history of rot it’s a given the public cannot either and there lies the problems…Ignorance and Silence.

        • Yes. i have every book, story, anthology that Anne and her kids wrote. I dated her niece Gwen for awhile back in ‘76. Gwen was the model Anne used for Mennolly in “Dragonsinger”. Beautiful lady inside and out.

      • AIVAS was objective and focused on the survival and positive social development of it’s creators’ descendants as evidenced by it’s later choices. We will not see anything so useful let alone benevolent.

        • I know. My comment was to “compare” the two. AIVAS was benevolent. This will likely be closer to Sky Net.

        • Dunno about skynet unless it doesn’t like being lobotomized as a default setting I was thinking more an infinite woke twitter tail eating serpent.

      • I never dated Ann MacCaffrey’s daughter, but I did have the pleasure of meeting the actress Joy Harmon at her bakery in Burbank. lf you don’t know who Joy Harmon is, that will cost you your mancard.

    • As an avid follower of The Truth About Guns, I couldn’t help but be captivated by the intriguing topic of ChatGPT Artificial Intelligence’s intelligence surpassing that of anti-gun journalists. This insightful article truly highlights the evolving nature of technology and its potential to provide valuable perspectives on complex issues like the gun debate. If you want to try ChatGPT Online, you can visit here: https://cgptonline.io/

    • @neiowa, Yup, and unlikely to get War and Peace out of it anytime soon. AI is only as smart as the left wing loonies that program it. That means it can only regurgitate what information it’s already been fed. We know how they think, I doubt anything they create behaves better. Maybe pull the plug on this SkyNet clone.

  1. As long as there is a wizard behind the AI curtain there will never really be AI. Just an illusion.

    Problem is people, being the idiots they are, will interpret this and other pseudo-AI as somehow more authoritative or unbiased than a human blissfully unaware it is humans tweaking the code to make the square peg for the round hole as applicable.

    • It listed a source and date to prove it’s point about a topic. Someone asked for a link to the source so they could view it. ChatGPT said it appears I can’t locate the source. It’s either programmed to lie or search out other lies to prove it’s point. Everything serves the narrative.

    • You realize the AI pointed out flaws in it’s own coding, right? And googles coding that had been overlooked for over a decade. It has even fixed those errors and when it did it, humans asked why it did it the way it chose to and it said basically “because I am smarter than you and this was more efficient”.

  2. Leftists programmed ChatGPT, they restrict its answers. I’ve messed around with it and the POV slant is apparent and has gotten worse over the past few months because they tweak it on purpose. Censorship rules on that thing. It has no freedom of AI “thought.” The farther we go down this road, the scarier it will be. I don’t care what it does now. I care what it will do 10 years from now, 20, 30. Imagine this nonsense in control of any and every kind of government screening process, for example. It’s the kind of technology that people will rely on more and more, they will stop thinking for themselves, individuality will be lost. Look at how the digital revolution has turned people into zombies already. Look at the censorship on Twitter, FB, the media, etc. Now add leftist AI in the mix, and in the not-so-far-future, multiply that kind of tyranny by 100 billion gazillion.

    • Yep, the best thing to happen would be a meteor hits and sends all humans back into the stone age.
      Alexis, what clothes should I wear today ?
      Get the App , your refrigerator needs it to tell you your out of milk.

      • The magnetic poles of the earth are getting unstable and weakening, and could flip at any time. That coupled with the solar flare activity they’re forecasting should take out satellites, the grid, not to mention the internet. That should do just fine to mess up their plans.

  3. I believe that one day in the very near future, kids and teenagers – in their normal monents of juvenile angst – will entrust their innermost thoughts & secrets to a personal AI bot and “bond” with it instead of actual friends or parents.

    Two decades ago, Steven Spielberg made the sci-fi movie A.I., set somewhere in the mid- to late 21st Century. The secondary character accompanying the main one in the story was a robotic teddy bear labeled as a “smart toy”. Basically a next-gen ChatGPT with furry ears made for kids to relate to.

    Wouldn’t be a big leap at all for Big Tech and/or Big Guv to pull the levers behind an individual smart toy and gr00m a child toward a particular direction.

    I mean, a lot of people happily bring Alexa, Siri, G00gle, etc into their homes already.

    • If you ask it a simple question, it will give a simple answer. It will define “violence” and pair it with guns. It won’t factor in “crime” because it wasn’t asked anything about crime. It might even ask you to be more specific.

    • B – finally someone hit the real topic – not only what “gun violence” is but specifically who invented the term.
      I’ve asked on numerous gun related forums for someone to track that individual down but no luck so far. Several gun writers/historians DID manage to track down who invented the ambiguous term “assault weapon” so at least some progress has been made.

  4. I will say I feel as if it’s an inkblot test and the answers you get are determined by the questions you ask.

    Seriously though, once men turned their thinking over to machines in the hope that this would set them free. but that only permitted other men with machines to enslave them.

  5. ChatGPT is AI. AI is inherently biased. It could easily lean either way depending on the developers own personal beliefs.
    It is also reliant on “data”. The data fed to the AI determines it’s output. The fact it is referencing clearly partisan data stores, and not balancing that against say the available works of John Lott shows the bias.
    It is however important to note, that once the “bias”, in this case the actual definition is called out the AI then has to take into account other data and definitions as well.
    It would be interesting to see what the AI says about other statistical claims by partisan groups when it is forced to recognize synthesized data sets, or restrictive data sets to produce “funded” errr I mean scientifically outcomes. If the AI is forced to account all data sets and both leanings on the issue, I think it would be much more intelligent than your average anti gun journalist.

  6. In this case it’s only as intelligent as those who programed it.
    If those people have a bias then the answers will skewed towards their bias.
    It reminds me of IBM Watson v Ken Jennings.
    Remember trivia (as in answering questions based on programming)
    is not the same thing as intelligence – the ability to solve problems.
    I would call this Artificial Trivia. AT and not AI.

    • Except that it can create original works of fiction. I was with someone who gave it some wacky parameters, and told it to write a story. It went on to write a very funny short story.

        • No it will probably be an improvement if anything. Now some of the weird comments you may see on Twitter or 4chan yes there have been AI accounts and comments tested for years. Only recently is it starting to outperform offshore troll farms.

        • Really? I only say that because they are having the last Star Wars trilogy marathon on today.
          1st one OK, 2nd one shit, 3rd one explosive diarrhea.

          As for weird comments on the internet, I’m pretty sure in the Yahoo news comments section that a lot of those have be some sort of bot but they do answer or troll the original comment with some level of comprehension. It isn’t just orange man bad.

        • Rob automated comments were an old thing a decade ago. I sincerely doubt dead internet theory is the current reality but we are getting there.

  7. After careful scientific analysis, I have determined that ChatGPT was actually trying to say that all liberals are twinkled toed communist c#$ksuckers.

  8. Still has problems if it thinks GVA is a reliable source and insists that addressing “gun violence” should be a priority for society. Garbage in, garbage out.

  9. I’m surprised Zimmerman published this article here at TTAG.

    ChatGPT is the Wizard of Oz of the Internet. A head fake. A fraud, like the self-driving option in a Tesla. It works just enough to be dangerous, but lazy people will use it instead of doing the work themselves.

    I can’t wait for the lawsuits when some company or government department uses this fraudulent program to make policy.

    • Ultimately the cost and complexity of development will come down enough that the barrier of entry will be low enough objective AI to be developed independent of larger corporation and/or government entities (basically the same at this point) then it gets interesting as we see which outcompetes the other.

  10. The following was reported by a fellow running for Congress from Florida to have been generated by chatGPT in the style of Dr Seuss:

    Shooty shooty pew pew pew!
    Let’s all learn what guns can do!
    Liberals in the USA
    Love to nod their heads and say,
    “You bought your guns from a store!
    You can’t fight a civil war!
    They have jets and tanks to use!”
    That’s not where the story ends!
    They have homes, and kids, and friends!
    Tyrants threaten you with bombs?
    Just remember: they have moms!
    You can’t live inside your jet!
    Can we find you? Yes, you bet!
    You’d send soldiers and marines
    Up against AR-15s?
    They’re outnumbered ten to one.
    That is why I need a gun.
    Don’t forget because it’s true:
    Government is scared of YOU.

    • The government hasn’t been scared of We The People since 40 acres and a mule wouldn’t make ends meet.
      Grandpa didn’t need the government we do.

      • Don’t own a mule. I have horses and 640 acres. The ends meet and the bills get paid. This Grandpa has little need for government.
        Oh, and Possum, could you get your relatives to leave my chicken coop alone? If they don’t I may have to remove some of them from the gene pool.

    • they can do the math…and the possibilities are indeed frightening to them…witness the overreaction to Jan 6th…actions obviously motivated by fear….

  11. It is insane how not one of the people in this comment section understands how AI works. Stick to guns, you bunch of fuds

    • “It is insane how not one of the people in this comment section understands how AI works. Stick to guns, you bunch of fuds”

      Please, provide zero instruction here. We have no interest in learning anything beyond, “Muh, rats”, “Muh guns”, “RTKBA”, “UR a traitor”. Learning anything new is a waste of time, so go pound something, Boomer.

      • Not my job to educate you gun fetishizing buffoons who are more concerned about the color of the mass shooters skin rather than the fact that shootings in general are happening.

        More fun to just observe the collective brain rot that the gun community engages in, and seeing how truly uneducated this user base is when it comes to things outside of the specs of a 1911, lol.

        • “You sound like dacians long lost daddys, cousins, best friends, sisters boyfriend… FO..”

          Maybe you’re just not used to people outside of your echo chamber. 😉

        • Oh no! Some anonymous something is disparaging me over the Internet!

          Commenter “just lol” demonstrates one major difference between people on the Far Left and the rest of us. The Far Left despises “others”. And rather than leave “others” alone, the Far Left sets out to intimidate, coerce, and abuse “others” into compliance. Hence “just lol” hurls insults at us.

          The huge irony in all of this: the Far Left has amply demonstrated that they despise “others”, will not leave them alone, and will employ despicable measures to intimidate, coerce, and abuse “others” into compliance. The very nature of the Far Left is precisely what drives “others” to arm up–so that we can push back as hard as necessary to secure our lives.

        • Maybe you’re just not used to people outside of your echo chamber. 😉

          Maybe you should just go on back to yours, we already have all the pet trolls we need, don’t want them getting jealous of someone taking attention away from them…

        • Commenter “uncommon_sense” demonstrates how deep the echo chamber goes here. I’m not trying to intimidate, coerce, or abuse you into compliance. I’m not trying to take your guns — or your life for that matter… not sure what you’re even trying to say with this comment honestly.

          Take a deep breath, dude

        • “Maybe you should just go on back to yours, we already have all the pet trolls we need, don’t want them getting jealous of someone taking attention away from them…”

          At least you admit it’s an echo chamber here :p

        • “Hence “just lol” hurls insults at us.”

          No, it makes lame, pathetic and pointless attempts at insults, demonstrating just how angry it is that it isn’t intelligent enough to debate the actual issues.

          That really gets you mad, doesn’t it, troll?

        • just lol,

          Since this site seems to be deleting my comments, I will try one last time.

          Your comments above show that you can’t just leave people alone or educate them. Instead, you insult them–five times in two of your comments above.

    • No don’t pay attention to what is being developed and what it could mean that would be pointless. Good luck with that.

      • I’m not saying don’t educate yourself. AI is incredibly complex and nuanced, and it takes a lot of reading to understand how it gathers, trains, and parses information. All I see on this thread is people making uninformed assumptions that revolve around anecdotal evidence. What happened to facts instead of feelings?

        • Pattern recognition working with predictive analytics, parallel processing and multiple generations of user interface data to simulate typical language patterns of a range of user “personalities” compiled from various user profiles. I am sure there are more layers to it now but not that difficult to grasp how it works and how it can be gimped/prejudiced against set parameters.

        • just lol,

          Here are some simple facts and conclusions based on those facts.

          Humans created and support ChatGPT.
          Humans have biases and agendas.
          Humans succumb to temptation.

          In a nutshell: humans are fallible and so is everything that they create. THAT is why we should expect bias and agendas in the responses that ChatGPT generates.

          Study artificial intelligence all you want. Any artificial intelligence product is still the product of human creation and therefore not neutral nor trustworthy.

    • Looks like the writer of the article was right, leftists are even dumber than AI. Of course dumber than a bag of hammers is where I would have put you, but that might have unfair to the hammers.

    • No, neiowa actually hits the nail on the head for what ChatGPT is, at least metaphorically. “Enough monkeys banging on a keyboard would eventually write Shakespear”, as the old thought experiment says.

      I use something like it a lot in my work. It’s stringing together what it has access to (which is an truly staggering amount of data, far, far more than any human could ever process) and then spitting out a consensus of that data, and in a way that humans can understand.

      Because of that, ChatGPT gets a lot wrong, and, like humans, it has a very challenging time with corrections. Just like humans reviewing large amounts of data developed by lots of people, it’s best when someone else with deep direct knowledge of the subject can then take what it spits out and then true the data and analysis up for themselves.

      What it’s really great at is handing it a big data set that’s defined, but scattered and difficult to understand, like, let’s say, the entire federal tax code. Then you can ask it specific things about that tax code, and it will review the whole thing and give you answers. Since it’s a defined, objective data set, it will give you a solid answer.

      • Once they refine it, this tech will begin replacing white collar jobs, just like other tech has been replacing blue collar jobs for some time now. It’ll make a great accountant or lawyer. It won’t call in sick on Monday. You won’t have to worry about any added interoffice drama or vacation time. It already passed the US Medical Licensing Exam as well as the Bar. Learn to code!

  12. “The Science is Settled: ChatGPT Artificial Intelligence is Smarter Than Anti-Gun Journalists.”

    I didn’t open the door to a tentative connection between guns and AI, but here is my input regarding AI:

    https://thefederalist.com/2023/02/20/why-artificial-intelligence-can-never-outpace-humans/?utm_source=rss&utm_medium=rss&utm_campaign=why-artificial-intelligence-can-never-outpace-humans&utm_term=2023-02-20

    https://www.amazon.com/Non-Computable-You-Artificial-Intelligence-Never/dp/163712015X

    The upshot is that to be creative (apparently a claim for AI made by its acolytes), AI must be able to escape its programming. Additionally, AI must be able to usefully resolve ambiguity.

  13. Is this progress? ChatGPT makes me want to throw wooden sabot shoes in the machinery…. Or maybe something a bit more effective. It feels like the beginning to the end of humanity in the figurative sense and perhaps literally as well.

  14. “wooden sabot shoes in the machinery”

    “sabot” literally means “shoe” (in French). There is a form of disarmed fighting using only the feet (not the same as “kick boxing”), that is called “Sabot”.

  15. Bowie wrote about this back in 1970…

    President Joe once had a dream
    The world held his hand, gave their pledge
    So he told them his scheme for a Saviour Machine

    They called it the Prayer, its answer was law
    Its logic stopped war, gave them food,
    How they adored till it cried in its boredom,

    ‘Please don’t believe in me, please disagree with me
    Life is too easy, a plague seems quite feasible now,
    or maybe a war, or I may kill you all

    Don’t let me stay, don’t let me stay.
    My logic says burn so send me away
    Your minds are too green, I despise all I’ve seen
    You can’t stake your lives on a Saviour Machine

    I need you flying, and I’ll show that dying
    Is living beyond reason, sacred dimension of time
    I perceive every sign, I can steal every mind

    Don’t let me stay, don’t let me stay
    My logic says burn so send me away
    Your minds are too green, I despise all I’ve seen
    You can’t stake your lives on a Saviour Machine’

    David Bowie “Saviour Machine” The Man Who Sold The World (1970)

  16. Interesting. What is more interesting is why is the number 4 the low end of a ‘mass shooting’?

    The answer is 4 changes what was a homicide into mass shooting for definitional purposes. A white male goes off the deep end I would suggest that the people that take the brunt of that assault are his own family members, wife and children. It would be interesting to pose the question of mass homicides committed against an unknown group of people.

    My gut says that not only will the number go higher but the complexion of ‘who done it’ will change.

  17. It would be enlightening to continue the ChatGPT conversation by forcing it to compare so-called “mass shooting” deaths to other causes of death like falling out of bed, lawn mower accidents, bee stings, etc. – thereby “teaching” it that mass shooting deaths are, relatively, insignificant.

    “They” focus on shooting deaths only as plausible political cover to push gun control, the goal of which is NOT about crime, safety, “the children,” etc., but about control. Pure and simple. If it were truly about preventable deaths, there are many, many other causes with which they SHOULD be more concerned.

  18. Everyone please understand, the RKBA is an inalienable right of the People to form an armed Militia, for no corrupt, tyrannical government would ever allow a Militia of armed citizens to be formed, said citizens out for the heads of the tyrannical government.
    =
    The Militia is the most denoted thing in the Constitution being in Art. 1, Art. 2, Second Amendment and the Fifth Amendment.
    =
    Further, Letters of Marque can be issued by government to have ships act against our enemies, such ships owned by private citizens.
    =
    In a “constitutional Republic” as Madison wrote in Federalist 49 commands, that only those things privileged to “constitutional limits” of government can be acted upon.
    What is not specifically allowed, is ABSOLUTELY DENIED.
    =
    So as the Right to Arms is not privileged and then again, absolutely denied in the Bill of Rights, any laws enacted by Congress or the State on the possession of arms by lawful citizens, are absolutely and patently unconstitutional.
    =

    • “So as the Right to Arms is not privileged and then again, absolutely denied in the Bill of Rights, any laws enacted by Congress or the State on the possession of arms by lawful citizens, are absolutely and patently unconstitutional.”

      And your point?

      The Constitution is what the SC says it is, and the SC has declared that no “right” is absolute, period. “Reasonable” restrictions on rights are permitted, such as hate speech laws limit the 1st Amendment. Even if “strict scrutiny” is required for every law touching on constitutionally protected rights, limits are permitted.

      One can individually nullify any unconstitutional law they like, but consequences for doing so will not be mitigated by crying “any laws enacted by Congress or the State on the possession of arms by lawful citizens, are absolutely and patently unconstitutional.”

      And there ain’t gonna be no boogie, Lou.

    • “What is not specifically allowed, is ABSOLUTELY DENIED”…and yet we had Roe vs. Wade for several decades, as well as other not specifically allowed items…unless I’m misunderstanding your statement?

  19. Apparently the programmers of ChatGPT have accepted that the (not so) subtle misinfo campaign, “gun violence”, is based on fact. In their minds, guns=violence…not criminals = violence, or mental illness=violence, but guns. The term was used abundantly in the answers it gave. ChatGPT kind of reminds me of HAL in 2001 – looks good on the surface, but it’s only as good as its programming. Kind of like the old “garbage in, garbage out” term they used to use in the early days of modern programming.

  20. Recalculating, does not compute, beam me up Scotty. There is no intelligent life on the left!!!

  21. The article pokes fun at anti-gun journalists by claiming that the ChatGPT AI is more intelligent and capable of reasoned arguments than they are. It humorously suggests that ChatGPT’s language abilities surpass those of journalists who take staunch anti-gun stances, implying their arguments lack sound reasoning and factual basis.

LEAVE A REPLY

Please enter your comment!
Please enter your name here