Twitter | Search | |
This is the legacy version of twitter.com. We will be shutting it down on 15 December 2020. Please switch to a supported browser or device. You can see a list of supported browsers in our Help Center.
Kaj Sotala
I keep seeing all kinds of crazy reports about people's experiences with GPT-3, so I figured that I'd collect a thread of them.
Reply Retweet Like More
Kaj Sotala Jul 15
Replying to @gwern
First, 's crazy collection of all kinds of prompts, with GPT-3 generating poetry, summarizing stories, rewriting things in different styles, and much much more.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Automatic code generation from natural language descriptions. "Give me a page with a table showing the GDP of different nations, and a red button."
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Taking a brief technical tweet about GPT-3 and expanding it to an essay which the author of the original tweet mostly endorses.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Acting as a more intense therapist than ELIZA ever was.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
On the side of "it does still have weaknesses too", figuring out how to trick it at the Turing Test.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Also, using the architecture behind it to generate images.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Redditor shares an "AI Dungeon" game played with the new GPT-3 -based "dragon model", involving a cohesive story generated in response to their actions, with only a little manual editing.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
The official announcement of the Dragon Model and examples. "The game invents a complex magic system and underlying theory behind why it works and describes the whole system as I read the book."
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Another example of automatically generated code, this time giving GPT-3 a bit of React code defining a component called "ThreeButtonComponent" or "HeaderComponent", and letting it write the rest.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
From a brief description, GPT-3 correctly generates an explanation indicating that it's a case of asthma, mentions a drug that's used to treat asthma, the type of receptor the drug works on, and which multiple-choice quiz question this indicates.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
GPT-3 tries to get a software job.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Translating natural language descriptions into shell commands, and vice versa.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @xuenay
Given a prompt with a few lines of dialogue, GPT-3 continues the story, incorporating details such as having a character make 1800s references after it was briefly mentioned that she's a nineteenth-century noblewoman.
Reply Retweet Like
Kaj Sotala Jul 15
Replying to @bioothod
. noted that not *all* of GPT-3's answers are that great. Fair. IMO, being able to produce good answers even *with* human selection is amazing by itself, and often it requires relatively little, but yeah it's not a general intelligence yet.
Reply Retweet Like
Kaj Sotala Jul 16
Replying to @xuenay
So far I had only been sharing other people's experiences, which did occasionally make me wonder whether they're all real. Now I tried out GPT-3 myself, to generate some space opera. And whoa. Yeah I believe that they're real now.
Reply Retweet Like
Kaj Sotala Jul 16
Replying to @xuenay
Reply Retweet Like
Kaj Sotala Jul 17
Replying to @xuenay
An earlier link pointed out that GPT-3 doesn't realize that nonsense questions are nonsense questions. But you can also just tell GPT-3 to treat nonsense as such.
Reply Retweet Like
Kaj Sotala Jul 17
Replying to @xuenay
Using GPT-3 to help you do gratitude journaling.
Reply Retweet Like
Kaj Sotala Jul 18
Replying to @xuenay
Source is an anonymous image board poster so could be fake, but: if you give an AI Dungeon character fake wolf ears and then ask her to explain formal logic to you, she may use the ears in her example. (source: )
Reply Retweet Like
Kaj Sotala Jul 18
Replying to @xuenay
Even after seeing all the other results, I honestly have difficulties believing that this one is real.
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
Starting to see lots of "GPT-3 is overhyped and not that smart" articles now. Sure it's not actually intelligent, but the fact that a non-intelligent thing can do so many things is still significant and it will have lots of applications.
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
Writer gives GPT-3 the beginning of an "How to recruit board members to your startup?" essay, thinks the second half is good enough to add into the essay.
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
Automatically creating designs from a verbal description.
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
Richard "GPT-3" Dawkins summarizes the central idea of his book, "The Selfish Gene". (Also GPT-generated conversations with other famous people in that thread.)
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
GPT-3 as a general question-answering system. (If I were to use that, I'd make sure to check those sources it gives, but of course even an answer that you have to verify from the source is better than nothing.)
Reply Retweet Like
Kaj Sotala Jul 20
Replying to @xuenay
Integrating it to Google Sheets: fill in surrounding cells, GPT-3 figures out how to autocomplete the contents of the next one.
Reply Retweet Like
Kaj Sotala Jul 21
Replying to @xuenay
Reply Retweet Like
Kaj Sotala Jul 21
Replying to @xuenay
Reply Retweet Like
Kaj Sotala Jul 25
Replying to @xuenay
A 200-page fantasy novel co-written with GPT-3.
Reply Retweet Like
Kaj Sotala Jul 26
Replying to @xuenay
Reply Retweet Like