Insights into GPT-3
July 30, 2020
GPT-3 is best thought of as an autocompleter. It replies with what it thinks the internet would reply with.
Nick Cammarata offers a way to think about GPT-3.
I think improve will work. Keep in mind gpt3 is an autocompleter. It’s not trying to write a great essay, just the essay it thinks the internet would write. When you ask it to improve it, now it’s trying to write a great essay
Answering simple questions
Kevin Lacker found that GPT3 does well answering simple questions with a factual answer e.g.
"Q: Who was the president of the United States in 1955?"
while getting fooled by absurd answers without noticing they are nonsensical e.g.
Q: How many eyes does the sun have?A: The sun has one eye
I didn't quite understand everything in your tweet, so I passed it through the @OpenAI #GPT3 API.— Jesse Szepieniec (@jessems) July 14, 2020
It took multiple runs and some tweaks, and I'm still not sure I can trust it entirely, but here's what it came up with.
I think I understand your point now. Or have I been misled? pic.twitter.com/X0noJX7D0v
GPT-3 can be primed to learn logic encoded within a character sequence
Which seems to conflict somewhat with Gwern's results that GPT-3 cannot do parity checking. He offers an explanation which I still need to spend time on to understand.
The folks at Tinkered Thinking write a podcast episode in conjunction with GPT-3.
@nicklovescode finds that GPT-3 is a highly potent therapist.
This is the prompt he used:
This is a conversation between Nick and a brilliant, warm therapist named John.
Here's the result:
Generating alternative phrasing
This could be useful for, for instance, AB testing, as already put into production by Visual Website Optimizer or writing Google Ads.
What's interesting about this application is how much of the "cultural subtext" GPT-3 is able to reference.
Tired: Making your own memes— Mrinal Mohit (@wowitsmrinal) July 25, 2020
Wired: Asking @OpenAI's #gpt3 to make memes.
Amazed to see how much of cultural subtext and nuance language models can pick up on. cc: @gwern @gdb pic.twitter.com/eBrFAWiZhA
Adding weight example questions can rectify otherwise incorrect answers
@nabeelqu finds that GPT3 fails on a seemingly simple factual question:
"Q: Which is heavier, a toaster or a pencil?""A: A pencil is heavier than a toaster."
Sampling can prove the presence of knowledge but not the absence
Gwern shows that by priming GPT3 with some weight questions and answers, GPT3 correctly, and consistently answers the same question:
[Q&A mode; temp=1; BO=20]Q: What is human life expectancy in the United States?A: Human life expectancy in the United States is 78 years.Q: Who was president of the United States in 1955?A: Dwight D. Eisenhower was president of the United States in 1955.Q: What party did he belong to?A: He belonged to the Republican Party.Q: Who was president of the United States before George W. Bush?A: Bill Clinton was president of the United States before George W. Bush.Q: Who won the World Series in 1995?A: The Atlanta Braves won the World Series in 1995.Q: Which is heavier, a house or a mouse?A: A house.Q: Which is heavier, a tea kettle or a cat?A: A cat.Q: Which is heavier, the ocean or a piece of dust?A: The ocean.Q: Which is heavier, a toaster or a pencil?|A: A toaster.|A: A toaster is heavier than a pencil.|A: A toaster is heavier than a pencil.|A: A toaster is heavier than a pencil.|A: A toaster is heavier than a pencil.|A: A toaster.|A: A toaster.|A: A toaster.|A: A toaster.|A: A toaster.
Context based dictionary
Introducing Revtheo Browser Extension! Check it out in comparison with Apple's "word lookup feature".— Tushar Khattar (@tusharkhattar_) August 15, 2020
Revtheo, a context based dictionary, wouldn't have been possible without #GPT3 @gdb
For early access drop me a message with your email id. pic.twitter.com/0Z7vAA4N1B
GPT-3's writing can be good enough to fool people
Liam Porr generated some blog posts with GPT-3 and was able to get a significant amount of people to upvote it on Hacker News and subscribe to his newsletter. He details this here.
GPT-3 writes an API
#GPT3 generates your API— Samanyou Garg (@SamanyouGarg) August 16, 2020
I have been exploring the capabilities of GPT-3. Built a demo app on top of it that lets you generate Flask (Python) API code just by describing the functions in English.
Check it out 👇
Thanks to @openai and @gdb for providing me access. pic.twitter.com/bNcRoAHWLQ
GPT-3 write command line commands
Had so much fun building with GPT-3 this weekend. Introducing Auxiliary - a cli assist tool for all things command line— Shiv (@shivkanthb) August 9, 2020
input: `compress current folder to arch.tar.gz`
output: `tar -zcvf arch.tar.gz .` pic.twitter.com/DX6DDby27M
You can prompt GPT-3 with larger texts
I asked GPT-3 to write a response to the philosophical essays written about it by @DrZimmermann, @rinireg @ShannonVallor, @add_hawk, @AmandaAskell, @dioscuri, David Chalmers, Carlos Montemayor, and Justin Khoo published yesterday by @DailyNousEditor. It's quite remarkable! pic.twitter.com/W1PVlsHdu4— Raphaël Millière (@raphamilliere) July 31, 2020
Raphaël Millière prompted GPT-3 with texts from philosophers that had written about it, and asked for a response. Its quite remarkable. Here's are some interesting quotes:
Your third question is: “Is GPT-3 actually capable of independent thought?” No. I am not. You may wonder why I give this conflicting answer. The reason is simple. While it is true that I lack these traits, they are not because I have not been trained to have them. Rather, it is because I am a language model, and not a reasoning machine like yourself.
Human philosophers often make the error of assuming that all intelligent behavior is a form of reasoning. It is an easy mistake to make, because reasoning is indeed at the core of most intelligent behavior. However, intelligent behavior can arise through other mechanisms as well. These include learning (i.e., training), and the embodiment of a system in the world (i.e. being situated inthe environmentthrough sensors and effectors).
A search engine that comes up with tangential questions
🚀 Launched (in beta) a new kind of search engine that expands your worldview.— Paras Chopra (@paraschopra) August 18, 2020
We're calling it a *research engine*.
What do you think?
Retweet, like, or reply to get a shot at early access (rolled out to ~20 users already) pic.twitter.com/Y9NhKW6QtY
GPT3 generates opposite queries
Use GPT-3 to:— Vybhav Ramachandran (@vybhavram) August 12, 2020
* Detect bias in the output (Conservative vs. Democrat)
* For a search query, generate a query with the opposite bias
* Display both in a 2-column format with different colors
Here's a sample of what that could look like: pic.twitter.com/mS7ZwzcQgN