Linden Lab's intention to use AI

[Gayngel]

Just read the year in review post and just want to say please please please Linden Labs do not relegate support to an AI chatbot! Every company that has done this has the worst customer service!

Edited December 31, 2023 by Gayngel

[Chip Craig]

Initial support via an AI chatbot is actually a good idea. Consider it to be just a replacement for a knowledge base - you say what it is wrong and it suggests known fixes.

Should the AI fail to resolve your issue then (depending upon your membership level) you still have access to various other support options. In theory, this should free up support staff to deal with proper support issues instead of "I can't be bothered to rtfm issues." In theory, providing it is limited to initial support only then it is good.

[Henri Beauchamp]

I see no problem with it, as long as the user (the ”resident”, in SL vocabulary) is properly (and prominently) made aware of the use of AI.

Also, there should be an option to opt out entirely from AI interactions (and that would be very wise from LL, given the upcoming regulations regarding AI and personal data).

Edited December 31, 2023 by Henri Beauchamp

[Gayngel]

2 hours ago, Chip Craig said:

Initial support via an AI chatbot is actually a good idea. Consider it to be just a replacement for a knowledge base - you say what it is wrong and it suggests known fixes.

Should the AI fail to resolve your issue then (depending upon your membership level) you still have access to various other support options. In theory, this should free up support staff to deal with proper support issues instead of "I can't be bothered to rtfm issues." In theory, providing it is limited to initial support only then it is good.

That is how it's supposed to work but on the ground it really doesn't. All the bot does is give you the run around trying a million fixes you've tried before and then tells you to e-mail support where you waste more time going back and forth when really the issue can be resolved or logged as a bug to be fixed in 5 minutes if you could just speak to a person in live chat or a phone call, its frustrating as hell dealing with support bots.

Edited December 31, 2023 by Gayngel

[Theresa Tennyson]

21 minutes ago, Gayngel said:

That is how it's supposed to work but on the ground it really doesn't. All the bot does is give you the run around trying a million fixes you've tried before and then tells you to e-mail support where you waste more time going back and forth when really the issue can be resolved or logged as a bug to be fixed in 5 minutes if you could just speak to a person in live chat or a phone call, its frustrating as hell dealing with support bots.

Yes, but you'll be on hold for an hour and a half waiting to talk to the live tech because there's a massive line of people in front of you asking questions that could be answered by an AI bot.

[Nalates Urriah]

AI support is not so much a good or bad thing. Good or bad will be determined by how the Lab uses it. If they train it well and use it to assist staff then I expect it to be a good thing. If they get sloppy in or shortcut the training and use it to reduce staff then it will be a problem for residents.

The state of most bots in SL is VERY primitive. And in general, most bot support anywhere has been primitive. The programming and computers driving them have been minimal. New chips for AI processing and massive data sets are changing how well bots do or do not work.

Interestingly, new AI bots can lie and make stuff up. This could make for some surprising answers from support. It would be like getting 'help' from someone in a 'Safe' Hub.

[Rowan Amore]

Aren't they already using it with Boxy?

[Claireschen Hesten]

I thought LL had some kind of AI already, the last time I had to engage with live chat i had to go through the Boxy bot. When the bot couldn't provide the answer a live chat agent stepped in and was most helpful