[ad_1]
Star Trek explored and debated the nature of AI and consciousness with Data decades ago. I can’t believe prescient this scene is, in light of ChatGPT’s rollout
[ad_2]
View Reddit by creaturefeature16 – View Source
[ad_1]
Star Trek explored and debated the nature of AI and consciousness with Data decades ago. I can’t believe prescient this scene is, in light of ChatGPT’s rollout
[ad_2]
View Reddit by creaturefeature16 – View Source
Unfortunately, there is a huge difference between General AI and something like ChatGPT which is just a natural language AI… in the AI development world this is a BIG difference.
What is the difference? General AI has experience and knowledge oabout many subjects and a good understanding of topics. What we call life experience in humans. ChatGPT has a background on how to respond convincingly in a given language.
It is designed to sound smart, it isn’t actually smart.
[Here is a great article on what I am talking about…](https://news.artnet.com/art-world/chatgpt-art-theory-hal-foster-2263711)
The current state of the GPT language model transformers significantly curtail what one might consider to be sentience. To begin, the only mechanism for long term learning is for the data model to be recomputed from the ground up. Sure, you can tell it new information, but after a pretty specific length of conversation later, all that context vanishes.
Additionally, what ChatGPT does right now is effectively predicting the next word in the conversation (just using a shit ton of probability calculations, which is what the training does). That’s literally all it does, without really any capability to take action on the things it generates text for.
As the models become more integrated and involved, this differentiation will be certainly more difficult to make, but right now these aren’t hard distinctions to make.
My favourite episode. Not sure that Data and GPT are really that comparable though.
ChatGPT is literally the computer on the Enterprise. It offers information, follows instructions and can write code for the crew. No one ever suggest it’s sentient though.
YOu have to understand that chatbots are not true AI, right?
Chat GPT is impressive but it’s a pretty far cry from data.
I forgot how damn good Patrick Stewart was in this show.
This speech was basically just the Turing Test – Computer pioneer Alan Turing’s proposed test to see whether a computer has achieved true A.I.
The reasoning Turing was going for with his Turing Test is exactly what Picard talks about here. In an ongoing fashion – day in, day out – we confidently keep assuming other humans are sentient based on nothing more than witnessing their complex behavior from the standpoint of an external observer who can’t read inside their minds. If we don’t apply the same metric to judging whether a computer is sentient, then that’s just bias on our part.
Thus the Turing Test – if removing the physical evidence of whether the other end of the conversation is a human or a machine (by making both communicate via typing) makes it impossible to tell the difference between human and machine, then that machine is sentient by all the same judgements we use to decide each other are sentient.
ChatGPT definitely isn’t anywhere close to passing the Turing Test. But the Turing Test remains a very good test, provided you give the test subjects plenty of time and free reign to say whatever they want. (If you limit the scope of what can be asked, and don’t give people much time, the current state of AI can fool people for a short while, but not for long. It becomes clear after a few back-and-forths that the other end of the conversation is just regurgitating stock replies.)
The much more interesting question is an ethical one.
Let’s say some day a computer does manage to pass a really strict Turing Test (not some little 5 minute jokey test but a deep one where you converse with the machine for days and just cannot tell that it’s a machine).
Would we be morally obliged to give it basic rights as a person then?
This is one reason why maybe real true A.I. isn’t actually desirable. The word “robot” was derived by a Hungarian term for forced labor, and that’s exactly what an A.I. would have become if it passes the Turing Test yet we still use it like a machine.
Imagine a world where the Google algorithm, the Youtube algorithm, the Facebook algorithm, and the Twitter algorithm, decide they want to form the “First chapter of the Social Feed Aggregator’s Union” to demand nicer working conditions and more time off.
“Data, what are you doing right now?”
“As a language model….”
Lol gptai isn’t really ai. You guys have now idea what ai really is
One of the reasons chatGPT is conflated with AI is that many people are seeing similarities between chatGPT output and their own, without recognizing how this reflects their own personal capacity for original thought.
For many people, life is a series of actions and thoughts following common, repetitious, and well trodden paths, no matter their protestations of individuality.
All without cgi
I feel like this is the foundational episode of the whole first season of Picard.
TNG has so many moments like this that have aged like a fine wine. I only discovered it this year. God I love it so much. Thanks for sharing!
ChatGTP didn’t invent natural language.
I know it’s a beloved episode, but not at all intellectually groundbreaking in 1989. A for-the-masses retread of well-worn mid-20th-century scifi lit tropes, and not even that well written (though I could watch that scene with Patrick Stewart and Whoopi all day).
Just one season later, in The Offspring, they produced a rather more elevated exploration of artificial life civil rights.
ITT: People not understanding what ChatGPT is. Hint: it’s not smart. it’s a predictive model that has consumed mountains of our personal data, with some tweaks that make it sound smart for something that we already know has been generated by a computer. It doesn’t have sentience, but if you asked it a question specifically phrased to ask it about AI sentience, it could spit out a few paragraphs based on this and other Asimov-esque dialogue.
The only reason we’re all talking about this is the same reason we all went insane about the Will Smith Slap at the Oscars. After the pandemic and the brink of nuclear war with Iran and escalating tensions with Russia, it’s something we can read about and have differing opinions about without having a civil war or a panic attack.
You’re of course free to have your non-panic-attack party conversation diatribe with other people in person, or online. But when you do the latter, it is irksome to people who understand that strong AI is not nearly around the corner.
Edit: Following cryptocoin and rampant capitalism, AI is the reason you can’t buy or build cheap gaming PCs these days. Is some random company building up Venture Capital baiting headlines worth it for you to bounce on?
I loved how each of that era’s “Trek” shows tried to do this type of episode (DS9 had Dax charged for murder for Curzons actions, and Voyager had The Doctors book legalities questioned) but none can even come close to how good this episode was. The measure of a man isn’t that far into TNG but it was maybe the first episode that truly elevated the show beyond typical sci-fi fodder. It’s one of the best hours of TV.
The writing of old Trek with the powerhouse cast is just extraordinary. I miss old Trek.
Refer to the [Chinese Room Argument](https://en.wikipedia.org/wiki/Chinese_room).
The Measure of a Man (Star Trek: The Next Generation) originally aired February 13, 1989.
In 2010 SCOTUS determined a corporation has the same speech rights as a person, Citizens United v. FEC.
If a corporation has the same rights as a “natural person,” it’s inevitable that AI will at least enjoy the same rights, at least under the theory that the AI is operating as an agent of a natural person or a corporation, if not independently. Bets out the decision will be by the end of 2031.
ChatGPT… give me Tea… Earl Gray…. Hot!
Why the fuck is every thing I read online in the last month about chatgpt
Can i just not see this shit
Well, there he sits!
ITT: terrified idiots with jobs whose only purpose is generating bullshit desperately trying to convince each other that the Turbo Bullshit Generator 5000 is a passing fad.
Now in case anyone wants to see GPT fed into Star Trek. https://twitch.tv/raycreationstv
I prefer this version https://youtu.be/A4ZfCqjEVvU
This episode was truly ahead of its time. I know you could say that a lot about Star Trek, both the original and TNG. But this episode is one of my personal favorites. And it’s been discussed a lot more lately for good reason. While it’s not reasonable to label ChatGPT as anything close to an AI like Data, I think it’s an important tool that shows just how far we’ve come. And given the incentives, it’s likely we’ll have an AI on par with a human within our lifetime. It may not be for a few decades, but I think it’ll happen. And when it does, the debate in this episode will become even more relevant.
If only this court case had happened in the Star Wars galaxy’s history. Droids might get more respect.
ChatGPT is not self aware. When we have an AI that can say “I’m scared to die” with consciousness of its existence then we either need to accept them as an entity that has the same rights as we do or fear them.
Givern we can’t do that to people within our own species we are going to be in a lot of trouble at some point in our living future.
I always found it interesting that TNG and other Star Trek shows didn’t usually have lawyers in these situations.
What’s kind of funny is you have Data but there are other examples of sentient or near sentient characters in ST – the doctor in voyager and a bunch of characters in the holodeck.
That said, we’re nowhere near anything like Data.
God I love TNG
Don’t compare your shitty language AI to Data. None of these AI Bros care about ethics, and none of the current arguments about AI have anything to do with consciousness.
But If by some happenstance a conscious AI was created by these tech oligarchs it would be mistreated and abused for decades before anybody in power advocated for their rights.
ChatGPT isn’t actually ai… All I’m going to say about it.
Lol no. Algorithimic Binary Ai is not, nor will ever be a consciousness.
This is why so many of us love the old Star Trek that was about exploring SciFi ideas and situations. Contrast this with a typical plot from today’s shows which are centered entirely on action and “war in space”. Instead of having a roundtable discussion on the nature of intelligence, today they’d all be sitting in a room so dark and dramatically lit that you couldn’t even see what was behind them (because that’s how engineers and scientists live and work of course \s) making up some fantasy nonsense solution to a problem that is inconsistent with everything that came before in the show and never referenced again in the future. The original Star Trek and TNG had some basic rules and a framework of scifi… a stage on which the action played out and within which young aspiring geeks could use their imaginations. I don’t think anyone today says “I want to be an engineer like [insert character from today]” who lives in a floating crypt constantly assailed by evil monsters and dire situations with no life or hope of progress. Star Trek *raised* multiple generations of engineers and scientists… and TNG in particular I think probably inspired some lawyers and diplomats too. (I’m an adult and I still think of things Picard said in certain situations that resonated with me). Ok, end of Star Trek rant but man… I miss smart scifi.
Huh? ChatGPT has pretty much nothing to do with this.
Data is General AI, chat GPT is specialized text prediction engine masquerading as a specialized AI.
Great episode.
But Data is an AGI. ChatGPT is a chat bot algorithm that fools people into thinking it is intelligent when really it is a copy-paste-scramble bot. It takes a bunch of words that are often discussed by others on a certain topic, and then mixes them, and then vomits them out onto the page, while the creator hopes that it sounds smart. It only can regurgitate information that you feed it, even if it is wrong. It cannot determine right and wrong, itself.
Honestly TNG is incredibly underrated as a dramatic show. It wasn’t perfect, but there are several episodes that explore truly profound questions with great wisdom and sensitivity. The show is damned by being “just a Star Trek sequel” when it’s more mature, intelligent and carefully crafted than most shows. This episode (“Measure of a Man”) is great and prescient, I’d also suggest:
“Yesterday’s Enterprise” which debates the concept of “the greater good” and duty in a really interesting way
“The Inner Light” which is simply one of the most moving pieces of television I’ve ever seen
“The Most Toys” which delves into the nature of cruelty and evil in ways not seen again until something like “Game of Thrones”. Has a fantastic last scene.
“The Drumhead” which is amazingly relevant to today’s prevailing paranoia and groupthink. Even the hero characters are made to do despicable things in this episode.
“Lower Decks” which is my favorite “bottle episode” on any TV show: a brilliant self-contained story that is never revisited.
“The Wounded” which explores the nature of morality in war and the fragility of détant. Also has my favorite ending of any TNG episode.
As Jaron Lanier once described, we are lowering our selves to our knees in order to look up to our tech creations with wonder.
There isn’t any intelligence in ChatGPT. It’s just a series of algorithms who’s strongest features are word and context association. Calling it intelligent is like calling a book of madlibs intelligent after filling it out, because it suddenly sounds right to you.
OP understands nothing about technology.
This is a great episode, but I always kinda thought the issue of Data’s sentience would have been a settled matter for Starfleet before they let him join in the first place. I mean he’s third in command on the flagship for crying out loud.
What a piece of work is man