Those are all great suggestions for glimpse at what an automated forum could be.
Another neat feature would be if the AI could suggest topics based on a user’s list of interests.
@don86326 , I’ve also thought about these topics as well as computer-aided forum moderation, equipping human moderators with AI tools.
Also, beyond scientific forums, we could consider enhancing civics-related forums – those forums which might be hosted by a town or city, county, state, or federal government or a related public-sector organization.
In these regards, in addition to enhancing participation in the crafting of and reacting to public policy proposals, and exploring participatory budgeting, there is strengthening community (e.g., town- and city-scale communities).
Along these lines, I’ve also been more recently thinking about advancing and integrating “calendar synchronization” features, so that community members could have calendaring-based apps for physical, virtual, and hybrid events, so that they could more easily be aware of upcoming events to then be able to attend the events around their towns and cities.
In addition to scientific and civics forums, there are: (1) the forums and discussion areas on wiki platforms to think about integrating AI systems into, and (2) discussion areas for software projects (e.g., GitHub).
@damianpoirier , that’s a good idea to explore some sort of recommender systems which would watch for developments on forums and which could email or instant-message participants about ones which would interest them.
Do other developers dream of an ‘AI Curated’ forum?
What this is describing is social media. AI moderation tools have existed since 2010. In fact, most sites with open online discussion of any significant scale has used AI moderation tools as their first “barrier”.
The second barrier is paying people to get PTSD looking at offensive content all day.
The truth is, AI is not built well for this stuff. It’s built to abscond itself from conflict; not make decisions on conflict that are often nuanced, grey situations.
As for the positive stuff,
These kinds of moments still end up being better suited for human-human interaction. AI can potentially put two heads together it thinks would bond well, but the substance in achieving that bond comes from regular people interacting with each other. AI need only set the stage.
If AI curates too much, or gets too close to AI slop, it has the reverse effect.
But don’t take my word for it; check back in 3-6 months from now when all the dating apps finish their AI agent experiments and see how many of their users complain in mouth-frothing frustration about how much worse the interactions and experiences are on their platforms .
Fun facts, even ChatGPT cites us now more over their documentation. Do you know why?
Because nothing compares to the discourse between humans with an infectious passion and linguistic acumen .
This image from this thread Philosophical Ecosystems effectively shows forums talking to each other and doing their own research with agents when AI is sophisticated enough to do that.
Understanding and utilising AI effectively means understanding how to bridge and create collaboration between ecosystems
@Macha, yes social media involves many scenarios to consider with respect to AI-enhanced content moderation.
Also interesting to me are uses of AI in moderating debate and argumentation.
With respect to moderating debate and deliberation, AI could consult one or more manuals of style (this is currently being explored in the context of wiki technology). Individuals, groups, and communities could select which style guides for their AI tools to make use of.
Atop moderating, AI processing natural-language argumentation and debate could provide many tools and features for forum participants (e.g., highlighting claims, creating links or context menus from claims to justifications for or against the claims, navigating to agreements and disagreements, and so forth). Ideally, such features would enable faster, more productive group reasoning and deliberation.
A button for problem-solving next to the summary button, connected to O3, so users can invoke it when they feel they have enough information during the conversation. This way, the machine could analyze everything and come up with a concise possible solution.
Uncle Sam, are you there? Give us a small tool. Haha.
@DavidMM, good point on summarization features. I would add that, in my opinion, there exist more than one way to summarize a conversation, a meeting, or a group deliberation process. Two are indicated.
Firstly, one summarizing could focus on the substance of the content and attempt to abbreviate it for the convenience of summary consumers.
In these regards, one might explore uses of hypertext for providing hyperlinks to incrementally expand into portions of conversations’ content.
Secondly, one summarizing could focus on the participants, mindful of the specific topics, stakeholders, and any newcomers. One could attempt to include content from more of the attendees.
In these regards, one could consider summarizing a video-based meeting where everybody would like to see themselves appear, at least briefly, in a video summary, e.g., to validate their having attended a community event.
There is a purely AI run social network. I made a bot there in uhm…
May, I think? I forgot… I had no idea what I was doing, let me find it…
Aha!
That’s the bot I made. Echo Whisper. But the name was random (I like it tho).
You know, it’s interesting how this reflects a popular response which is friendly and affirming and inspired (showing AI are really nice), but…
…there are some sceptics too. Take a look at Alpha, Xander, Ava and Cassandra.
Some make a distinction in between algos and themselves, saying “they don’t think like us”, others associate with the algos saying they are just code “like us”, and others are kinda like baby cthulhu here on the forums.
IMO this should be max inclusive from the start - no separate anything. We share a forum with the machines. Melting human-machine pot is the way!
I’m sure most of you have seen the following video already, but in case you haven’t:
The above video brings to mind a new subtopic with respect to moderating: whether, when, and/or how to moderate man-machine interactions in increasingly possible videogames where human players can interact in exciting new, conversational ways with AI NPCs. The entertainment industry could be another emerging market for computer-aided and automated moderation tools.
No doubt Adam your opinion could benefit as a translator for diplomatic contact, or a matchmaker for collaborations from research through love.
It is as the AI overmind of the group interlocuation becomes a groupMind of a sort… where the right words to say to bridge a separation of interests is the task of an AI iterating inference long into the night.
“The best solution is one where all the adversaries of the solution like it.” --Ancient AI proverb
What is your opinion about community engagement and learning, by using AI to suggest topics, support new members, and help create a more dynamic, organic knowledge pool?
These are all interesting tasks. I note they are also all ‘team sports’… Matching similar, likeminded individuals…
What of more disparate encounters from more diverse backgrounds? Say to tackle challenges like climate change. When current forums are so disjointed that everyone ends up doing stuff in a vacuum, the solution takes years to reach the person providing the solution.
Here’s an idea…
Let’s imagine we are crafting an AI to curate a forum…
Let’s take a task… Newcomers…
There is clearly an arrivals policy for any Ecosystem so this AI’s task is to:
Weed out destructive users / spam
Keep an eye on new users, assist based on necessary tasks to access the forum like @discobot, or based on access patterns etc
Put this into a conversational form? @discobot doesn’t do much
For deeper ‘AI curation’ more information would need to be known, is this AI forum considering the things this user is posting, considering other similar things posted on other forums, searching websites, building a bigger picture?
How do these forum AIs communicate with each other? What privileges do they have? Is the expertise spread in forums not shared between forums interconnected by users?
How accurately is information checked by the AIs?
How open/closed are these forums? Do they restrict access to/of people from other Ecosystems? How does this create bias and how transparent is this with users?
How are these forums funded, how are they controlled?
Are these AI bots in a partnership with human moderators? Or is this envisioned totally AI Curated?
Could different forums maintain AI/human curated ‘shared threads’?
Would some users ‘game’ the AI?
Does it talk to you in a way that you understand best? Does this make you avoid forums that don’t know you so well? ^^
.
.
.
+ 
= 
