Embracing The Potential Of Artificial Intelligence

Hey there, MusicBrainz fam! Long time no chat! :notes:

First off, I want to give a big shoutout to the staff, moderators, and everyone behind the scenes for keeping this community alive and kicking. You all are rockstars! :metal:

Now, I have a confession to make. I’ve been lurking around these parts more than actively participating lately. Life threw a few curveballs my way, and I haven’t been able to log in and contribute as much as I’d like. My bad.

I also want to apologize if I’ve made any flippant remarks in the past. When I’m not in a good mental space, my words can sometimes come across as not-so-great. I’m sorry if I offended anyone—I genuinely didn’t mean to. :pray:

But hey, let’s talk about the future, shall we? The potential role of AI and AGI in lightening the load of manual data inputting sounds pretty darn exciting, doesn’t it? Imagine the time we could save! While we’re not there just yet, it’s definitely something to keep an eye on.

To all you amazing folks who’ve been keeping the music flowing and contributing day in and day out, thank you! Your dedication keeps this community thriving, and I’m truly inspired by each and every one of you. I’m itching to catch up on all the awesome things you’ve been up to!

So, let’s kick off a fresh start together. I may have been a silent lurker, but I’m ready to jump back in, support one another, and embrace the future of MusicBrainz. And if I slip up again, feel free to give me a friendly nudge—I’m committed to being a positive presence in this community once more. :blush:

Please see existing “AI” topics

It seems the community isn’t particularly thrilled about automated data input.

1 Like

Welcome back DerekFerrick!

I’m excited to see AI/machine learning put to good use! But I think we should treat obviously bot-written posts (like some of your recent ones, this one included) as we would spam.

People using bots as a tool to help get their own ideas across or to help them get cross language barriers is wonderful. But people can visit chatGPT themselves if they want its unfiltered ‘thoughts’ or answers. We don’t need it presented in the forums as if it was a users own writing/thoughts. Particularly if it’s sharing incorrect information.

I believe this is already covered by point 9 in the MetaBrainz code of conduct.

The current community manager @reosarevok can clarify the official MeB/forum stance.

Perhaps you can explain your use of AI in these posts @DerekFerric - is it helping you cross some kind of communication barrier? If it’s for fun then that’s cool, but I think we don’t want a forum that’s just chatGPT talking to itself, in the long-term :wink:

9 Likes

Seeing a ChatGPT post is the same as someone posting a Google Search. Likely to be close, but not an actual accurate answer based on real knowledge.

Personally I find it pretty rude when people reply to me using LLM-based text generation. It gives the appearance of not even taking the time to type something out yourself and presuming that I wouldn’t have thought to search for an answer using these tools (or some other ones) myself before posting.

2 Likes

in the long term u get a personalized AI that is trained on you and will eventually vote for you removing the need to representative democracy and go for full direct AI voting where the personal AI will vote for every single law, norm and regulation

I appreciate the link. I must say I’m quite surprised to hear (at least some of) the community doesn’t share my optimism. i guess I’ll get my coat!
Seriously though, if work is being done to complete the database, why does it matter if we use GPT or anything else to improve our communication during the background here in the forum? I got a GCSE in English so don’t technically need it, but I’m also using a clunky keyboard with a million unread emails to get through on my current machine and would like to have no typos, not insult anybody accidentally with my attempts at humour due to my allergy to the work of N-Sync, for example and do a better job. Would people be upset if I was using Grammarly? How is this any different?
Bizarre. However, I’ll respect the majority and or bosses’ wishes. It just mean I can’t work so quickly. To hide my post in a thread about tackling the issue of repeating tedious tasks using scripts seems to contradict everything you guys and gals stand for. Is it because you’re regretting having been so generous with your time over the years now it’s emerging that you maybe could have just waited for AI, as I myself feel having spent years archiving stuff myself and fixing the utter chaos and mess of other people’s attempts at analog-digital transcribing etc etc? I can totally relate. I might have got married and become a father if I hadn’t been doing my work so far in the metadata sorting of a public collection a relatively few amount of people in the retired rave scene are now undertaking before all the cassettes in our possession dissolve or deteriorate beyond repair. You want the database to be perfect, but not us. I’ll hold my tongue so as not to offend - but consider what Spock or Data from Star Trek would say!

Absolutely, re; language barrier. I asked GPT to rewrite my post to you in the other thread to make me sound less of a berk! :smiley: I should have known that I couldn’t pass as a human that easily. LMAO!

1 Like

But it was my thoughts, not Google’s. I just used a translator.

Time is money though. I don’t have as long on the clock as perhaps you have and I want to finish my job before I die.

Is this place some sort of competition I’m not aware of? You’re not making sense to me.

If you don’t have the time to write a post, the best choice is just not to post.

4 Likes

Personally it is the inaccurate side of ChatGPT that concerns me. Too many errors are churned out. And as it reads grammatically correct the replies can get trusted too much.

Another technical forum I am on flat out banned it due to it being used by some people to respond to support requests. The dataset it was using was two years out of date and lead to some very bad answers due to this. Instead of assisting, it was giving very bad advice and wasting time for the person who asked.

I understand you wanting to use it to replace your words, but some of the posts you made just read odd and it clearly wasn’t you. It had translated it a long way from what felt like your words. The post at the top of this thread reads like a bad salesman. To me it just reads odd and generic :robot:.

I am someone who often gets his words very tied up on here and have got totally misunderstood on occasions. To such an extent I actually am literally scared to post some days. But I would still rather take a breath and try again than let a machine mangle my words even further. If I can’t judge when I get it wrong myself, it would be even harder for me to judge when the machine gets it wronger. :smiley:

I’d rather talk to you - warts n all - than try and converse with a machine. I wouldn’t want to see a forum get to a point of just machines talking to machines. I like humans too much - our faults are what makes us us. :grin:

8 Likes

I think I just came across the first user apparently making edit notes with ChatGPT or the like. They are extremely lengthy and not entirely factually correct either:

I hope this letter finds you in good health and high spirits. I am writing to propose a solution for addressing the issue of false or fake credits on MusicBrainz, with the aim of improving the accuracy and reliability of the platform’s credit information. As a passionate supporter and user of MusicBrainz, I believe that implementing certain measures will greatly enhance the integrity of the database and maintain its status as a valuable resource for music enthusiasts worldwide.

Problem Statement:
Currently, MusicBrainz faces a challenge in dealing with false or misleading credits attached to music releases. These inaccuracies can undermine the credibility of the platform and create confusion among users. It is crucial to address this issue promptly to preserve the integrity of the database and ensure that the correct information is available to the community.

Proposal:

Enhanced Verification Process:

Establish a dedicated team within the MusicBrainz community to focus on the verification of credits and ensure their accuracy.
Implement a systematic process to verify credits through reliable sources, such as official artist websites, liner notes, and reputable music databases.
Encourage the community members, including artists, music producers, and industry professionals, to actively participate in the verification process by submitting credible sources of information.
User-Reported Inaccuracies:

Develop a user-friendly interface on the MusicBrainz website and mobile app that allows users to report potential inaccuracies in credits.
Introduce a review system where reported credits are thoroughly investigated by the verification team, taking into account the evidence provided by the users.
Regularly communicate with users who report inaccuracies, keeping them informed about the progress of their reports and the actions taken.
Revision and Correction Mechanism:

Enable a revision and correction mechanism where trusted users or moderators can propose changes to credits that have been identified as false or misleading.
Establish a transparent voting system to ensure that proposed changes are thoroughly reviewed by the community before being implemented.
Allow users to provide supporting evidence for their proposed changes to facilitate the review process.
Clear Documentation and Guidelines:

Develop comprehensive guidelines and documentation that outline the criteria for verifying credits and the types of sources considered acceptable.
Make these guidelines readily accessible to the MusicBrainz community, providing clear instructions on how to contribute to the verification process effectively.
Education and Community Engagement:

Conduct awareness campaigns to educate users about the importance of accurate credits and how they contribute to the overall quality of the platform.
Organize regular webinars, forums, or Q&A sessions where community members can ask questions, share experiences, and learn from each other regarding the verification process.
Expected Outcomes:

By implementing the proposed measures, MusicBrainz can achieve the following outcomes:

Enhanced Accuracy: The overall accuracy of credit information on MusicBrainz will improve, ensuring that users can rely on the platform as a trusted resource for music information.
Increased Transparency: Users will have more visibility into the verification process, promoting a sense of trust and community engagement.
Reduced False or Fake Credits: The number of false or misleading credits on the platform will decrease significantly, minimizing confusion and misinformation.
Stronger Community Participation: The proposed measures will encourage active participation from the MusicBrainz community, fostering a collaborative environment focused on maintaining the quality of credit information.
Enhanced Reputation: By addressing the issue of false credits, MusicBrainz will reinforce its reputation as a reliable

If you really want to try interacting with that user PM me or use the edit search. Though it looks more like vandalism. (already reported for “spam”)

7 Likes

100% spam - sticking a “proposal” into an edit note shows lack of Intelligence to start with.

This kind of ChatGPT abuse is going to make a mess of many decent forums and databases. It is this kind of abuse that really worries me about the damage this kind of technology can do. Drowning quality sites under automated mess.

What worries me even more is the kind of person who thought that kind of post is “helpful” or “intelligent”.

3 Likes