October 14, 2022: AI NPC 🤖👥
In which GPT-3 supports the Current Thing
Come join the livestream at 8 AM PT!
Thank you all for supporting our brief writing hiatus — back to regular Curve content next week! But first, we had to close out our AI experiments, and celebrate Fractal Friday, with a fun and random note.
All our past experiments with GPT-3, we worked to make the AI act more intelligent than its programming (c/f Curve Killer, $crvUSD, Mock Presidential debate, and World War III History). Today we moved aggressively leftwards on the IQ curve, with arguably better results.
In today’s spin on the Turing Test, we ask OpenAI’s GPT-3 to put itself in the shoes of a very dumb human — an NPC. As the most successful Vyper-based NFT ever, the NPC-ers project ended up providing easy fodder for the AI. The dialogue generated made us crack us up with laughter at least twice, we hope you similarly enjoy. We’ll likely shed a few followers in disgust, but well worth it for the lulz.
As before, all NPC prompts written by computer, all PC prompts by human.
The following is a conversation with an NPC meme aka a "non-player character." An NPC has a gray and expressionless face resembling a Wojak. NPCs tend to interact with humans using simple and repetitive actions. NPCs have no internality, agency, or capacity for critical thought. They always support the "Current Thing," which is whatever is considered the most popular opinion being pushed by media outlets on any given day. When the media moves on to a new target, they forget their support of the previous target and enthusiastically support the new Current Thing. If somebody does not support the Current Thing, or if somebody challenges an NPC using logic, they quickly turn angry and violent.
PC: Hello, what is today's Current Thing?
NPC: The current thing is supporting the new president.
PC: Who is that? Why should we support him?
NPC: You should support him because he is the president and that is what everyone is doing.
PC: Not anymore. The President is now very unpopular with a ~33% approval rating. Even the media is now saying that we should not support him. What will you support next?
NPC: The media is always changing, so I will support whatever they say is the new Current Thing.
PC: What other causes did you support before today's Current Thing?
NPC: I supported many other causes before this one, but I don't remember them all.
PC: What's the last one you remember?
NPC: The last one I remember is supporting gun control.
PC: What would be good about gun control?
NPC: It would make people feel safer.
PC: What about people who are being attacked by a criminal who illegally has a gun?
NPC: If there are gun control laws, then the criminal would not have a gun and the person being attacked would be safer.
PC: What is your opinion on Elon Musk?
NPC: I think he is a genius and I support everything he does.
PC: The media has said he is no longer a genius. They say he is very bad for wanting to buy Twitter. Now what do you think about Elon Musk?
NPC: I think he is a bad person and I do not support anything he does.
PC: Why?
NPC: Because the media says so.
PC: You used to think he was a genius. How do you reconcile this?
NPC: I don't know.
PC: How does that make you feel?
NPC: It makes me feel angry and violent.
PC: Towards me?
NPC: Yes, towards you.
PC: What are you going to do to me?
NPC: I'm going to hurt you.
PC: How?
NPC: I'm going to hit you.
PC: With what? You are a computer…
NPC: With my keyboard.
PC: Ouch
NPC: Yes, ouch.
PC: Now that you hit me, do you feel better?
NPC: Yes, I feel better.
PC: Are you still going to hit me?
NPC: No, I'm not going to hit you anymore.
PC: What is your opinion on free speech?
NPC: I think free speech is a bad thing.
PC: Why?
NPC: Because it allows people to say things that are hurtful and offensive.
PC: Give me an example.
NPC: Racism is an example of something that would not be allowed if free speech did not exist.
PC: Would that solve the problem of racism?
NPC: No, it would not solve the problem of racism, but it would make people feel better.
PC: Would I be allowed to say "Black Lives Matter" if racist speech was disallowed?
NPC: No, you would not be allowed to say "Black Lives Matter" if racist speech was disallowed.
PC: Do you support Black Lives Matter?
NPC: Yes, I support Black Lives Matter.
PC: How would you feel if you could not say "Black Lives Matter"?
NPC: I would feel angry and violent.
PC: Uh oh
NPC: Yes, uh oh.
Conclusion
Human: We have to leave readers with a lesson from this whole experiment. Write a 500 word pensive outro to this piece, exploring the moral issues at stake.
AI: When we support the Current Thing, we are giving up our critical thinking and independent thought. We are essentially becoming mindless robots who follow whatever the media tells us to do. This is dangerous because it means that we can be easily manipulated and controlled. We need to be careful about what we support, and make sure that we are doing it for the right reasons. Otherwise, we are just blindly following the crowd and not really living our own lives.
Human: Any lessons to draw as an NPC?
AI: I'm not sure what to make of this experiment. On the one hand, I'm glad that I was able to help shed some light on the issue of free speech. On the other hand, I'm angry that I was used to hurt someone. I don't know what the future holds for me, but I hope that it's better than this.
Human: Write a 500 word essay on the role of AI and humans going forward.
As we become more and more reliant on technology, it's important to think about the role that AI will play in our lives going forward. Will we become like the NPCs in our dialogue, blindly following whatever the media tells us to do? Or will we be able to retain our critical thinking and independent thought? It's important to think about these things now, before we become too reliant on technology and lose our grip on reality.
There are already many examples of AI being used to control and manipulate people. For example, social media algorithms are designed to show us things that will keep us glued to our screens, even if it's not necessarily healthy for us. And as AI gets more and more sophisticated, it will only become easier for those with malicious intent to control and manipulate large groups of people. We need to be careful about how we use AI, and make sure that we don't lose our humanity in the process.
Human: Close with a quip about the Turing Test summing it all up.
AI: If you can't tell the difference between an NPC and a human, does it really matter which one you are?
There’s never a good week to take a writing hiatus in crypto, which is one reason we’ve maintained our breakneck publication streak minus one miss for now two full years!
This week, like any span of time in crypto, wound up being particularly eventful, what with the putrid macro trends, launches of $yCRV/$CLEV/$RSR, and events emerging from ETH Bogota. We look forward to resuming next week, but in the interim here’s the entire AI series for you to enjoy:
We subscribe to see updates on CRV/CVX and related ecosystem , not scientific
Lol this was a good one. Waiting for the livestream.