We’ve compiled a super quick survey (5 multiple-choice questions + 2 optional open-ended ones) to understand the interest and challenges around Generative AI, Large Language Models, and related topics within the Julia community.
The goal is to gather insights on:
Interest levels in this exploding field
Main hurdles for builders and users
Cool use cases you’d like to share
It’ll take you literally 1 minute to complete, so why not do it right now?
Your responses will help us shape the Julia Gen AI ecosystem better for everyone. Don’t miss this opportunity to have your voice heard!
We’ll close the survey in a week from now (3rd July).
As a general comment…in addition to the unclear phrasing @jar1 mentioned, I’d say several questions in this survey seem to take as given that the respondent is interested in generative AI.
For instance, question 2 does not give an option for “I don’t use these tools.” Question 4 is phrased as if everyone wants to use gen AI in Julia and are only being held back by challenges, and again is missing a “no interest” option. And question 6 assumes the respondent has a favorite use case for gen AI (compare “What is your favorite,” with a more neutral phrasing such as “Do you have a favorite”).
Not everyone is interested in gen AI, and these question phrasings might give you a not-fully-representative view of the Julia community’s interest in/needs around these tools.
All good points!
Q2 and Q4 were not designed for negative responses. I’ve changed that now.
Q6 was optional, so I think it’s fair to skip it if you don’t have an answer. I’ve made that explicit in the description.
Sorry about the confusion! Given that the tagline was “help us shape the Julia Gen AI ecosystem”, I assumed that only people interested in GenAI would open the survey / want to respond to it.
Yes, I’ve added a welcome message that clarifies that upfront “This survey assumes some interest in GenAI and Julia.”, so people not interested at all don’t waste their time.
If you only survey people who already think this stuff is good enough to use, you’ll get a slightly biased view of the challenges ahead.
For instance, I don’t use LLMs when I write open source code because I cannot guarantee that the results aren’t stolen from a repo with a closed license.
Absolutely. That’s why the phrasing used was “some interest in GenAI and Julia”. People have many interests and we do not need to collect responses from people who don’t care at all.
For instance, I don’t use LLMs when I write open-source code because I cannot guarantee that the results aren’t stolen from a repo with a closed license.
That’s a good example of a challenge and it can always be added as a comment in Q7 (open-ended one). I wanted to make the survey quick, so definitely lost some precision as a result.
As a side note, your concern is valid for any code. Humans make a lot of mistakes too.
Julia Gen AI Survey 
The goal is to gather insights on:
It’ll take you literally 1 minute to complete, so why not do it right now?
Take the Survey: JuliaGenAI - Community Survey 2024