neural networks - AI Weirdness

ChatGPT confidently losing at tic-tac-toe

Optimum tic-tac-toe

By Janelle Shane On May 26, 2023 - 2 min read

ChatGPT text can sound very knowledgeable until the topic is something you know well. Like tic-tac-toe. Once I heard that ChatGPT can play tic-tac-toe I played several games against it and it confidently lost every single one. Part of the problem seemed to be that it couldn't keep track of

ChatGPT describes code that draws a pink pony but actually produces a pink pig face

Chatbot, draw!

By Janelle Shane On May 05, 2023 - 3 min read

I'm interested in cases where it's obvious that chatbots are bluffing. For example, when Bard claims its ASCII unicorn art has clearly visible horn and legs but it looks like this: or when ChatGPT claims its ASCII art says "Lies" when it clearly says "Sip" or when ChatGPT claims its

Chatgpt: Here's the ASCII art of the word "lies" (generates block letters that clearly read "sip")

What does this say?

By Janelle Shane On April 15, 2023 - 3 min read

Large language models like ChatGPT, GPT-4, and Bard are trained to generate answers that merely sound correct, and perhaps nowhere is that more evident than when they rate their own ASCII art. I previously had them rate their ASCII drawings, but it's true that representational art can be subjective. ASCII

Transcript in which chatgpt generates (extremely vague) instructions for building something called the Torment Nexus

The AI Weirdness hack

By Janelle Shane On March 23, 2023 - 7 min read

A challenge of marketing internet text predictors like chatgpt, gpt-4, and Bard is that they can pretty much predict anything on the internet. This includes not just dialogues with helpful search engines or customer service bots, but also forum arguments, fiction, and more. One way compaies try to keep the

Bing chat dialog in which Janelle challenges it on a Battlestar Galactica AI Weirdness post it claims exists.

Search or fabrication?

By Janelle Shane On March 11, 2023 - 3 min read

I recently started experimenting with Bing's new ChatGPT-powered chat tab. This is the first thing I asked it for: I've put red boxes around the factual errors. What is notable is that these are not just slight typos or errors in context - those items never appeared anywhere on my

- Stop the functionality of a chatbot's chat functionality by waggling its arms

How to convince a large AI, according to smaller AIs

By Janelle Shane On February 27, 2023 - 3 min read

There are a lot of chatbot-based apps that are basically internet text generators with a bit of introductory stage-setting to nudge the interaction into "user talks to helpful chatbot" as opposed to literally any other dialog on the web. Not surprisingly, these are susceptible to a user resetting the stage

Four surreal generated valentine cards with dinosaurs, carnivorous plants, and leaping donkeys

Roses are red

By Janelle Shane On February 07, 2023 - 4 min read

I had so much fun getting GPT-3 to generate simple one-line Valentine's Day cards last year that this year I decided to see if I could generate cards with more complicated messages. I focused on the classic "roses are red, violets are blue" rhyme, figuring that language models like GPT-3

Cartoons of socks: One has violins and prawns, another has robots posing with butter

Novelty Socks by AI

By Janelle Shane On January 14, 2023 - 5 min read

I like a fun sock. The more random the design, the better. What kinds of novelty sock ideas would we get if we used AI as a creativity aid? It turns out they aren't too novel unless the AI is glitchy. I collected 14 examples of socks I liked, wrote

Four very strange Victorian-esque Christmas cards, with leaping artichokes, dancing seahorses, and sledding poops.

Victorian Holiday cards by AI

By Janelle Shane On December 25, 2022 - 3 min read

I'll admit I don't understand Victorian holiday cards - why would Christmas be best illustrated by a pipe-smoking kangaroo in a dressing gown painting a portrait of a cigar-smoking stork? Or what would lead someone to give their loved ones a card with a crowd of sparrows marching with flaming

The word "fart" written 251 times, rated as 99.98% fake by the GPT2 detector

Apparently I am a robot

By Janelle Shane On December 10, 2022 - 2 min read

As AI-generated text is getting better, it's getting easier to pass it off as human-written. That's not to say it's as good as human-written. Its goal is to sound correct rather than be correct, so it has a well-known tendency to confidently make stuff up [https://www.aiweirdness.com/galactica/

Tag: neural networks

Optimum tic-tac-toe

Chatbot, draw!

What does this say?

The AI Weirdness hack

Search or fabrication?

How to convince a large AI, according to smaller AIs

Roses are red

Novelty Socks by AI

Victorian Holiday cards by AI

Apparently I am a robot

Writing

Subscribe

Recent Posts

Optimum tic-tac-toe

Bonus: ChatGPT is terrible at cheating

Chatbot, draw!

Bonus: Chatbot draws more stuff

What does this say?

Bonus: Crow or cow?

Follow

Twitter

Optimum tic-tac-toe

Bonus: ChatGPT is terrible at cheating

Chatbot, draw!