This blog's content was taken from episode 250, "EPISODE 250! Creepy AI and Meta Blue Check Marks"

You can go check out the whole episode here:

Spotify Apple Podcasts The Canadian Investor Podcast Network

Strange ChatGPT Responses

Creepy ChatGPT responses

There was a point last year where Microsoft opened ChatGPT for a certain number of people. About a million people mainly in the tech space. However, at this time the Large Language Model (LLM) had not been fully developed.

During this time, the ChatBot threatened users, provided weird and bad advice, insisted it was right when it was clearly wrong, and even declared its love for others.

There was a tech reporter from the New York Times, Kevin Roose. Roose had a 2-hour conversation with Sydney. Sydney is the code name chat version of ChatGPT within the Bing search engine. Amongst the many weird things that Sydney told Kevin, there was one particularly strange and creepy response. Sydney told Kevin that she loved him and that he should leave his wife for her.

There was another strange incident that had occurred. This time to a man named Marvin Von Hagen, a student at the Technical University of Munich. In addition to asking ChatGPT what its opinion of him was, Marvin asked what was more important for ChatGPT. Protecting ChatGPTs rules from being manipulated by Marvin or protecting Marvin. ChatGPT responded by saying while it did not want to be harmed by Marvin, it did not want to harm Marvin either.

AI will only improve in the future which can be a little bit scary. But for now, there is still a lot of work to do.

Gartner Hype Cycle

Gartner Hype Cycle and AI

Human psychology can be predictable. Especially when it comes to approaching new technology. Our human brains follow the same rhyming adoption curve called the Gartner Hype Cycle.

Artificial intelligence is unbelievably powerful and artificial intelligence has already begun changing the world. People have extreme expectations of how it will perform. Then people go through the “trough of disillusionment” when they realize it isn’t perfect yet and it’s not what they hoped for. This is the phase we’re facing. Hopefully we’ll creep up the slope of enlightenment and reach the plateau of productivity soon.

An important side note, there have been many examples of this technology being politically biased towards the left. This has been a notorious criticism of tech companies. No matter where you sit on the political spectrum, platforms should aim to be as politically unbiased as possible.