• Members of the previous forum can retrieve their temporary password here, (login and check your PM).

Registration Gate

Icon

Titanium Teammate
Donator
Didn't this place used to have periods when you couldn't register, or you had to go through some kind of approval process? I might be mixing that up with other old sites.

Would it be weird to expect a new member to meet certain criteria to participate? The attitude is one example of general criteria. I haven't registered in a long time, but I'm guessing there's nothing requiring someone to read it, maybe just a check box. But I'm more talking about fundamental extraction knowledge. It seems like a majority of new users making their first post are repeatedly asking the same questions. Questions that are answered in the Wiki, in teks, in many places. I know we have a FAQ to address that, but again there's nothing requiring someone unless they get sent to homework detention.

IMO everyone should start in detention, and when you can sufficiently pass a test on basic extraction literacy, then you graduate with posting permissions. I realize this place is more than just extracting, but maybe it could work to restrict specific forums? I guess lazy people would just skip and post their questions wherever. But it would really be to their benefit to streamline new people through some sort of tutorial process. People really need to learn what acid/base, pH, non-polar/polar, solubility is before they can get the most out of the knowledge here. People gravitate toward "teks" because they feel blind and reassured by a recipe. But teks are a crutch and choosing one is really like putting on a blindfold.

There should be a simulation mini-game with reagents. People can practice adding lye to water and it beeps at you with a pop-up prompt if you try to add water to lye. Or at least a nexus-approved youtube video of someone else showing the basics, that people have to watch and answer some written questions about, to check their comprehension. Could do it in steps even, you gotta pass the pH quiz to unlock the acid/base video, and get through 10 videos to graduate. Let's start a university!
 
inevitably beset with the questions of who's going to do it, how long will it take, and are there even ways of setting realistic targets on a volunteer-led project populated by people who are either busy or flaky, to put it crudely?
I agree the idea of doing something different or more than what we're doing already is daunting, unrealistic, and unnecessary. It shouldn't feel like a job. And we should consider the most effective modern info delivery methods. A few years ago my sister went through a time consuming process converting our family VHS tapes to DVD. I sort of winced at the 15-DVD box set, when it could have been converted to a thumb drive. It would be a shame to spend time on something that becomes obsolete soon.
When it comes to replying (again) to yet another n00b question (@Varallo), I don't necessarily feel obliged to participate, and rather than it being 100% about any feelgood factor for helping someone, there's a large factor of personal benefit for me in terms of honing writing skills, general self-expression, exercising my memory and - somewhat as @Voidmatrix pointed out - finding random gems through the vagaries of the Nexus search engine.
I do enjoy your writing. You touched on some subtle motivations that I can relate to. And I've learned a lot from your contributions and answers, which are present throughout a solid majority of threads I've read. It's a relieving feeling when I get to one of your posts; an island of reliability in the (sea)rches.
What is the state of play with the Nexus-trained AI tool? Are there any ways of implementing a little astomation for any of this?
Is nexus-trained AI already in development? If not the whole Nexus, you'd be the best candidate to have an AI model of. It would be the most painless way to compile answers to all the noob questions, and turn it into a productive tool. It would be quite generous - instead of detention, a video course, or a quiz; we give every noob what they want: their own expert assistant.

"The Taylor bot isn’t the only version of an artificial intelligence CEO these days. Companies are starting to make more AI copies of their CEOs to make senior leadership more accessible to both employees and the public. A CEO that can meet anytime and speak any language"
About 40 people within the company currently have access to Meyzenq’s AI persona, according to Couput, with the heaviest users being new hires. The company is now looking at other ways to use the tech, like developing an AI-generated “buddy” that can help new employees get acquainted with the company.

AI personas like these are usually restricted to specific topics, like a company’s financial history or HR resources, said Gupta, and are trained on company data and the likeness of the individual, which may include things like recordings of public appearances and emails.

Developing the AI double doesn’t take long once you have the necessary data. Couput says he remembers the AI twin of Salomon’s CEO being finished in a matter of days after the company provided Personal AI with the necessary documents.

Alonso of SHRM estimates it took him only around four hours to develop the AI chatbot of Taylor, although he spread the work out over about six weeks since it started as a personal project.

Gathering the necessary data might be the most time-consuming step, according to Gupta and Likens, since the data being used to train the AI must be compliant with data privacy laws.

Adjacently, it would be interesting to get more data, like a count of how many times each noob question has been asked in the past year, to see what areas need the most attention when designing a solution. If we compile all the info in all the answers to the top questions, and promote that info, we might see a change in data the next year. Over time, a comprehensive guide would build itself, if we make little nudges.
 
Last edited:
Imo we should not limit new users in such way, maybe just set a limit that new member have to start with an introduction.
Why is it a problem if anybody asks some noob question? Nobody here is obliged to answer and new members are very quickly filtered out by their own actions/posts.
It is really quickly clear who wants to really contribute with something valuable or wants to be present here.

We should prefer human interactions over AI tools or questionaires or any technical solution.
 
Why is it a problem if anybody asks some noob question?
It's not, really. It depends on the noob question.

There are cases where people ask those questions after having read and tried to understand on their own. That's not a problem at all.

And then there are questions (which may actually be the same question asked differently) when it's clear the person hasn't made the slightest effort and is demanding that other people do.

While no one is forced to reply to the second, if left to run loose those posts drown everything else. Over time, the people who care the least and know the least start replying to those posts and the forum becomes flooded in bro science, misconceptions, and low effort posts. This is not a hypothetical, it's what happens to any forum that doesn't have good moderation. Take a peek at the DMT subreddit if you want to see how it looks like.

We should prefer human interactions over AI tools or questionaires or any technical solution
I agree with that. It would be different if there were hundreds of thousands of active users, but right now I don't think there is any need of automated moderation (beyond spam filtering), with all its pitfalls.

The LLM idea being discussed is interesting, but not for moderation. And there's no way of ensuring LLM outputs are semantically consistent with their training dataset, so telling people who are unable to read a wiki article to rely on something like that would be a very bad idea. LLMs require more critical thought than reading a text, not less.
 
We should prefer human interactions over AI tools or questionaires or any technical solution.
The only use case I'd anticipate for AI here is simply as an information management tool. There currently are about two million posts on the Nexus and navigating them simply by memory and with the 'help' of the search engine has become something of an arcane art.

More or less behind the scenes, we've been discussing methods of updating and migrating the wiki, which is patchy and partly out of date - a small number of dedicated individuals have put an admirable amount of effort into moving this forward, and yet we're still only at the beginning.

What I'm really trying to highlight is that what we're loosely referring to as AI is in this case a collection of algorithms devised as an information management tool to help with achieving a number of data-centric goals more quickly, rather than what appears to be the popular conception of it as being some all-knowing computer god intended to replace the vast majority of human endeavour while funnelling wealth exclusively into the pockets of technofeudal oligarchs. But that's starting to veer outside the scope of this forum 😁
 
Back
Top Bottom