Do we have any AI hobbyist among us
- Thread starter ukgold
- Start date
It's important to remember one basic thing about AI: it has no skin in the game. So, the name "AI" is a bit of a misnomer. AI is fundamentally very stupid because it suffers no consequences for being wrong. It also receives no benefit for being right. It's a zombie. Someone for whom there are consequences has to check its results.
I suspect there are marketing reasons behind the current AI boomlet. The basic premises of the tech haven't changed. The capacities are where they were. Brain research has revealed that modern super-sophisticated neural nets are laughably less sophisticated and complex than brains. And anything for which there is an on/off switch will always be uninterested in consequences. Which kind of guts any resemblance to "intelligence."
I suspect there are marketing reasons behind the current AI boomlet. The basic premises of the tech haven't changed. The capacities are where they were. Brain research has revealed that modern super-sophisticated neural nets are laughably less sophisticated and complex than brains. And anything for which there is an on/off switch will always be uninterested in consequences. Which kind of guts any resemblance to "intelligence."
I find it interesting as a person who enjoys creating content.the possibilities of creating animation is neat. Imagine we could see Rupp and Secretariat reaching us about Kentucky history or whatever.
I've used ChatGPT and Bard some and what I found interesting is that the answers sometimes have sort of a human quality to them as opposed to mechanical. For instance in one response it said " . . .by the way just as a reminder. . . " and
". . . don't conclude from this that X will do what you want, it may very well be that Y is the better choice for you.. ."
Those are more reflecting of an entity with conscience awareness as opposed to a machine producing text. I made me feel as if I should thank it for it's response. Nothing like Alexa or Siri
". . . don't conclude from this that X will do what you want, it may very well be that Y is the better choice for you.. ."
Those are more reflecting of an entity with conscience awareness as opposed to a machine producing text. I made me feel as if I should thank it for it's response. Nothing like Alexa or Siri
But isn't that the nature of how it was trained? It was taught how we talk and responds in ways we would like? The reason I don't feel like I should thank it is because it feels nothing either way.
Definitely some useful tools especially for content creation. And a total game changer for foreign scammers so watch yourselves.
Yes. It's trained to by whatever it was fed. It will use that data set and spit out combinations of words in the most likely order. So it 1) is only as smart as the data used to train it and 2) isn't really solving problems, just arranging words which is a minor but important distinction.
The bigger question in all this is: which companies allowed these huge AIs to feed/train off their data? It had to be Facebook, Microsoft, Google, etc. Nothing else had data sets large enough to train anything as sophisticated as the latest iterations. That needs cleaned out imo so everyone has full disclosure on what these companies are doing and how they benefit.
The next concern is long term effects on society. Entry level jobs are already being automated, especially anything data entry related. That cuts out training grounds for employees looking to advance. Next will be basic coding jobs and so on. You're going to have increased unemployment completely divorced from any economic factors and companies margins are going to grow to ridiculous unfathomable sizes.
Experts are projecting 73 million jobs will be lost to automation by 2030. Robots can perform routine manufacturing jobs, and jobs in food service. Autonomous vehicles will replace truck drivers, and cab/Uber drivers.
In my opinion this will create enormous stress on the economy and require some major revamping. While the efficiency and productivity increases will be beneficial, with a vast number of unemployed people
there is inadequate demand to support the economy notwithstanding the social consequences.
If you look how ai is being marketed by IBM and other large tech companies, they are pushing it as a way for your workers to be more productive. In reality, it’s about having fewer workers. But I think if things start getting bad (mass layoffs) you’ll see a tremendous backlash. And I think these tech companies know it too.
I think you'll see some staggering eff ups caused by over-reliance on AI (like the Jefferson County busing fiasco) that will create a distinct lack of trust in the whole damn thing and forestall it having too dramatic effect on society.
That or we will all just be construction workers and therapists in the future, I guess.
That or we will all just be construction workers and therapists in the future, I guess.
That's exactly how it's marketed to the public. To their prospective clients it's absolutely marketed towards reducing employees and reducing labor costs.
By the time the public figures out they were had, it will be much too late
In my opinion therapy is very formulaic. Like if you have OCD there’s a series of steps you go through.
You are looking at this all wrong. AI doesn’t replace a brain. Can’t. Never meant to. AI merely integrates human language, big data, and organizational skills. This would make it good at spatial tasks, data and studies integration, broad sweep data analysis for things like should I launch my nukes. The second you ask it for a moral judgement, you stop using AI and start getting a developer’s political views. At this point you just have yet another talking head. So a question like analyze the top 10 drivers and risk mitigations for lung cancers in every ethnic group is a fair question for a mature AI A question like when should a lung cancer patient be allowed to die is a stupid question for AI. That’s a morals personal choice type question. That’s not stupid AI. That’s a dumbass asking questions.
Now if one day we ask a question and the machine responds, I will need 72 geographically diverse redundant power feeds to answer that question, pull the plug. Kill it.
Lastly, if you are a middle manager, modestly skilled code writer, robotics designer, industrial artist, etc, you should start considering Plan B. You may need another source of income. Musk and Hawking warned us about this. Nobody listened.
I'll tell you what should be illegalized right now: AI phone answering systems. They don't work and they drive people insane.
I was always surprised at the gun store owners response when asked if he had a phased plasma rifle in a 40 watt range. " Hey, pal just what you see". That's it?