| General > General Technical Chat |
| Should the EEVblog Forum go private to prevent the robot uprising? |
| << < (6/9) > >> |
| coppercone2:
and I think maybe, if it can help, its not much different then turning off the lights of a city to reduce attack until some progress is made. It is a situation that is ongoing... and nothing is inevitable imo, I don't want to fall into dogma of AI being inevitable, maybe its not, to the disappointment of the financial backers. Dave jones makes a fair point that it is a time sensitive defensive situation. |
| ogden:
--- Quote from: tggzzz on March 15, 2023, 09:25:57 pm ---How do you determine that a human poster here is a sentient being? That's not as easy as I would have hoped. --- End quote --- This is exactly my point - that security/authenticity of current internet is insufficient. I am very concerned that literally anybody can get "unfiltered" advice. |
| fourfathom:
--- Quote from: ogden on March 15, 2023, 09:48:11 pm --- --- Quote from: tggzzz on March 15, 2023, 09:25:57 pm ---How do you determine that a human poster here is a sentient being? That's not as easy as I would have hoped. --- End quote --- This is exactly my point - that security/authenticity of current internet is insufficient. I am very concerned that literally anybody can get "unfiltered" advice. --- End quote --- Are you very concerned that literally anybody can get bad advice from a human posting here? It happens. What we do here is debate and explain, and digress and show off, but eventually the bad advice is replaced by good advice. The same thing is going to happen with AI-generated advice. |
| Dr. Frank:
I'm an absolute fan of 'Blade Runner'. If there would be a liability for an identification of AI created posts, no problem. Otherwise - 'you'll never know' - just kill them. Frank Btw.: Would a private forum prevent AI from entering, anyhow? I already have difficulties with some human users here, to distinguish real electronic experience from pure baloney, like simply replicating threads/knowledge from other users, evidently w/o any own experimental background. |
| RoGeorge:
I think the forum should stay visible. For example, I've found out about the EEVblog forum after Google Search popped a few very good threads from EEVblog, and after heaving the nice surprise to not be coerced to register for a click on a JPG and a 20kB download. Nobody knows yet how to deal with ChatGPT bots. :-// Maybe add some more human-verification tests for new accounts, or do like other forums where the first 10 posts or the first 3 days (whichever comes first) will become visible only after a moderator approval. Though, this would probably need much more than 3 moderators to cover 24/7 human approval for newcomer's posts. Responsible moderators are hard to find. My advice would be to make generic Moderator1 .. ModeratorN accounts, and keep those separated from the current username, so the moderators can come and go without altering their presence as normal users. Otherwise, in time they'll get hard fillings. Most of all, tell the new mods that they are to guard against vandalism only, i.e. against chat bots, and not to establish the technical correctness of a post. If a moderator thinks it should be 2k2 and not 1k2, then use the normal username to comment about technical solutions. Mods accounts must by used for administrative tasks only. |
| Navigation |
| Message Index |
| Next page |
| Previous page |