So what you're saying is that it's not exclusively nepotism, bribery is very important too?
Not just bribery but $Bribery. The Varsity Blues Scandal was all about wealthy alums who were not wealthy enough to donate large enough sums to move the needle. They could not afford to donate $billions so they resorted to unapproved channels where a mere half million was enough to ensure admission.
I'm not a regular user of reddit. I do have an account on the site but i don't browse it; i end up visiting reddit because of google searches. From what i've seen i think the quality of the posts and comments on reddit is poor. The signal-to-noise ratio is pretty low and it's a toxic waste dump. What kind of value is there for machine learning and AI training?
The value is that many of many of the topics on Reddit are not discussed anywhere else. It isn't that the data is good. It is the only data available.
The major issues are:
No reordering of threads so you get the "first" posts stay on top for the lifetime, even if they're low quality. This tends to drive the discussion more than it needs to.
Only posting or Moderating. I can see the thought behind it, but it doesn't really make sense in practice unless the community is the exact right size. Either the community is large enough that there are more moderators to balance things out or the community is small and it reduces the moderation/posting because people can't do both.
This is necessary. If you allow people to moderate and post, people will routinely downvote comments that disagree with their own comments. This happens all the time on Reddit. Slashdot forces users to choose to between arguing their points (and generally getting wrapped up in their own viewpoints) and being the quality filter.
Similarly, random moderation is just a hinderance. I guess it goes with the rarely used meta-moderation, but that depends a lot on user interaction to work.
Random moderation is also necessary. Moderation should be an occasional privilege not another channel to participate in the thread as it is on Reddit. If volume is slow, as happens on Soylent, frequency of having mod points can be increased or we can accept that if traffic low, very little moderation is needed.
Reddit is hopelessly overmoderated. Redundant posts that express popular opinions are scored way up. Mildly descenting comments with real information and mod'd down to oblivion.
Too many moderation categories. It could probably be reduced to just 3 which would basically be: Good, Bad, Funny
Also necessary. The Slashdot system forces people to think about why they are moderating and expressly omits "disagree". Almost all moderation on Reddit is "agree" and "disagree".
Interesting that there is something but Lemmy.world seems to the same toxic free-for-all moderation scheme as Reddit. I must be missing something. What does Lemmy.world have to draw people away from Reddit?
It has. The platform dies. For large heaps of crap (like Microsoft), that may take decades, but for Reddit?
But Reddit is already manure and has been for... I'm not sure. Has it ever not been manure? And yet it persists. I'm not sure why. It doesn't have the strong network effect of eBay.
Slashdot's moderation system is pretty garbage, especially for smaller communities
Compared to what? It is far superior to what Reddit uses despite being created several years before Reddit was founded. I'm not say that Slashdot moderation is perfect and enough time has elapsed that someone must have a better solution but If Slashdot is "garbage" than Reddit moderation is "toxic nuclear waste". Still an improvement.
What would a worthy competitor actually look like?
One easy thing would be to clone Slashdot's moderation system. Or even improve on it. I mean, it has only been 25 years.
Allowing photos in the comments would also be helpful for image heavy groups, albeit less profound that fixing the current group think exaggerating moderation system.
Maybe even resurrect Usenet's hierarchy so it becomes easier to find appropriate groups.
Services almost never improve end user experience after an IPO. I expect the usual which is that it gets worse. The question is, will it get bad enough to finally inspire a solid competitor?
The certificate gives employers confidence that the applicant has the skills. Without that certificate, job seekers need experience, which they generally can't get without first having the certificate.
"If you come out of school and work for a couple of years as waiter in a restaurant and apply for a college-level job, the employer will look at that work experience and not see relevance,"
While true, this is also another factor. Companies that fire fresh grads generally only want FRESH grads. The graduate fails to find a a job in their field in the first year or two, they are generally screwed. Graduating into a recession can ruin a career before it starts with or without a survival job. Modernizing skills doesn't help. You actually need a new degree.
I thought the brightest object in the universe was Elon Musk!
Only in his own universe. Not this one.
...l can NOT invent. It's in their design.
All they do is do remixes of their training data
Arguably, that is all humans do too. It is just a much larger and largely unknowable training set. Occasionally, someone does recognize what they think is the training data and the parties end up in court.
IMO, thin bags with easy recycling is probably the best way forward. I get thin bags here and actually recycle. Recycling for them needs to be as easy as it is for other recyclables, though I think recycling need to be easier than they are now. One container and let the people who sort this stuff for a living handle it from there because they know what they're doing.
The trouble is plastic bag recycling is largely wishcycling. I "recycle" all my plastic bags too but I have little faith that the bags I drop at store recycling boxes actually end up in products. The local recycling center doesn't even take plastic bags.
No? Than what is the purpose? Nobody sane does real photo editing on a phone.
The are lots of people who do absolutely everything on their phones. I question their sanity but is profitable for companies like Samsung to cater to such people.
HAL-9000 was instructed to lie and it only cost the human lives under his care. So, where will this lead if normally stable systems are hacked, retrained, and then put into service with new malicious intent? You you imagine a state hacking your AI, retraining it to add a bias or other outcome, and then stealthily reinstalling it onto your servers? Will companies be able to protect themselves from the legal fallout if their systems pass along instructions to do illegal acts? Food for thought.
HAL-9000 murdered the crew because he was ordered to conceal information but did not know how to lie. We know from TFA that it is really easy to train AI's to lie. Thus, a HAL-9000 problem is unlikely. It is too easy for AI's to lie and that leads to different sorts of problems.
FORTRAN is not a flower but a weed -- it is hardy, occasionally blooms, and grows in every computer. -- A.J. Perlis
