Reading_Traveller
Apprentice
I meant if they make something very similar to CUDA. For example how matlab and octave are very similar. Then what I wrote might be the case. But in any case, we know that NVIDIA is the clear winner here.
Meet ChatGPT (Artificial Intelligence), good alternative to Google?
- Thread starter lockhrt999
- Start date
Human communication, language usage, etc. will always evolve especially when they feel their communication and thoughts are being analyzed and exploited by by AI. There will also be competition between AI systems, so will have to keep on evolving to that.
Don't think it is going to be that simple.
Partially reminding of this: https://www.usatoday.com/story/news...erent-shut-down-creating-language/8040006002/
But some tasks can easily be handled, which it is already doing well and progressing...
In my opinion, AI will influence the human communication. We can already see that with Microsoft Editor. With the help of AI everyone will be able to speak Shashi Tharoor.
Those bots were built for a purpose. They will not work if you give them a different purpose. Those were created for negotiation. But if you ask them to create music, they aren't going to be of any help.
What I meant was, AI is now transitioning into AGI. Artificial General Intelligence. Essentially, like a human mind, AGI can discuss problems with itself and develop a solution. Current AI is like a teenager. Yeah, it's sharp as a tack but it's still dependant upon the world to provide it. AGI, on the other hand, is a full grown adult. An adult can think for itself. It can go out, take up a job, build its own home, take a loan, build a business etc. AGI won't have to rely on what humans are talking on the internet anymore.
There's a famous old experiment in this field. We know IBM's watson was the first AI who beat top ace human chess grandmasters. Watson was trained on the millions of historical chess data. Then came google and it's AI. Google didn't train it like the Watson. Google told it the basic rules of the chess and asked it to brute force learn the game, all by itself. There were a dozen matches played between Google's bot and Watson. And, the Watson never got to win.
AGI is like looking at Piano and then become the next Mozzart. You don't need to have access to a trainer or other people's music to sample. It's not required. When you are the Einstein, whom are you going to ask for help? You'll have to come up with something yourself.
Long but excellent read -
direct.mit.edu
Large Language Models and the Reverse Turing Test
Abstract. Large language models (LLMs) have been transformative. They are pretrained foundational models that are self-supervised and can be adapted with fine-tuning to a wide range of natural language tasks, each of which previously would have required a separate network model. This is one step...
direct.mit.edu
RaceGurram
Discoverer
Its natural tendency of Human to resist any change. Its one of the core principles of how we evolved.
1. Pre-Covid to Covid transition, everyone hated working from home, isolated
2. Post Covid, its same again. We hate to leave home and go to office.
The transition out of human comfort zone is always resisted. We resisted Industrial Revolution, and even computers when they introduced. We thought all humans will be replaced by Machines. There is nothing wrong about it, lot of routine mundane tasks will be replaced by machines.
The only difference, we have internet in hands to freely talk, make sensations, argue stuff now where as earlier it was normal chit chat between bunch of people
On the contrary, Humans do evolve and leverage the knowledge to become super intelligent. And people who can't get the same, will become replaceable materials.
1. Pre-Covid to Covid transition, everyone hated working from home, isolated
2. Post Covid, its same again. We hate to leave home and go to office.
The transition out of human comfort zone is always resisted. We resisted Industrial Revolution, and even computers when they introduced. We thought all humans will be replaced by Machines. There is nothing wrong about it, lot of routine mundane tasks will be replaced by machines.
The only difference, we have internet in hands to freely talk, make sensations, argue stuff now where as earlier it was normal chit chat between bunch of people
On the contrary, Humans do evolve and leverage the knowledge to become super intelligent. And people who can't get the same, will become replaceable materials.
Ai making people dumb.
I can see users asking how to do this?, how to do that? more often.
You should use your brain, read what it is saying, finding difficult terms meaning, learning more by trial and error.
When you write or type code, data or anything on your own instead of copy pasting, you understand and learn better.
I can see users asking how to do this?, how to do that? more often.
You should use your brain, read what it is saying, finding difficult terms meaning, learning more by trial and error.
When you write or type code, data or anything on your own instead of copy pasting, you understand and learn better.
Companies around the world have been outsourcing white collar jobs to India because it's cheap. Now, AI will do many of these jobs for even cheaper. I don't see why AI won't cause massive lay off in India and recession. Prepare yourself.
Kaleen Bhaiya
Innovator
As the saying goes, "ChatGPT won't take away your job per se, but those who know how to exploit ChatGPT, definitely will!!"
There are limits to what current generative AI can do, limit as in serious limits but it can be a very handy assistant to knowledge workers. Only real threat as of now is to customer care exces. Giving answers to problems which have been solved before is one thing, solving novel problems is another. ChatGPT and likes are phenomenal piece of tech but they are far from solving new problems which people solve on daily basis at workplaces.
On side note I believe someone somewhere right now who understands the generative AI and its limitations very well is working on something groundbreaking and is beyond use cases people like us can think like writing poems or leetcode problems.
On side note I believe someone somewhere right now who understands the generative AI and its limitations very well is working on something groundbreaking and is beyond use cases people like us can think like writing poems or leetcode problems.
blr_p
Oracle
Like which jobs. Show me the money
People said when mobile photography got good that wedding photographers would be out of work. No end of reviewers tried to show this or that phone as good as an SLR. Did it happen? no
When computer-controlled printing got started in the 80s it did put typesetters and engineers who had to configure printers manually out of a job. A new breed of printers that could understand postscript came out and the printing business was never the same again.
Has machine translation put translators or interpreters out of business? No, still required
I tried to translate a Russian document about their nuclear doctrine and I thought the translation was really good. Still going to need human oversight if it's to be used for anything official or commercial for that matter. The point is it's going to take a while for people to trust this tech before they let people go
This thread is full of nothing but fear-mongering and parroting of hype
Nice one. Jobs said the PC was like a bicycle for your mind. Improves productivity and that accelerated when those computers got networked and further still when those computers got miniaturised. Were people laid off in this process, no, economies grew worldwide and still continue to.
With each new iteration of tech, some jobs go away and new ones are created because productivity went up and that created a demand for something else. Nobody seems to be talking about that second bit.
When they wanted to computerise govt, there were protests by govt staff worldwide who feared the computer would take their job away. Instead, it created a demand for people who could write apps that went on for over four decades to the point there were not enough in the West and they had to import them from the East.
I've seen how well AI works in user communities like Google's G+. When the posts by the mods get deleted for spamming
Mods post more often, because they do the system decided they were spammers and removed their posts
I hear AI will be a real gamechanger in the military
AI’s Inhuman Advantage - War on the Rocks
When an AI fighter pilot beat an experienced human pilot 15-0 in the Defense Advanced Research Projects Agency’s AlphaDogfight competition, it didn’t just
warontherocks.com
AI pilot beats one with experience 15-0. It costs tens of $millions to train a fighter pilot.
So how soon is any air force willing to put one of these gadgets in their even more expensive jets? any takers out there in the real world
It will likely start with drones. Japanese got a 6G project on the anvil
heh, if humans are the ultimate arbitrators and they don't understand the solution provided they won't use it. Faster reflexes is fine though
Last edited:
Kaleen Bhaiya
Innovator
I guess you guys aren't following what's happening in AI industry. Current AI is showing the sparks of AGI.
Like I said previously, AI = teenager which goes to college to train. AGI = fully grown adult which doesn't need training on past data to solve unsolved problems. Equating AGI to just another calculator or printer isn't correct. Calculator, printer etc tools are made for humans. AGI (when it's ready) is human. It's made to replace humans. It's a tool for sure yeah. But it's the best suited for CEO and top management.
And I'm not fear mongering. I was the first to jump on the AI train. I have been using and paying github copilot for almost a year now, GPT4, signed up for its plugins etc. I'll soon start using AWS's copilot replacement which was recently launched. I have been closely following what's happening in this industry. I love it.
In my recent thread I asked something that I could gift to a 5 year old. My first thought was to print a similar book myself. I pitched the idea to GPT4, it created a nice story in the style of Roald Dahl. I asked its image generative AI to draw cartoon in the style of Sir Quentin Blake. Beautiful job here as well. Only problem was that cartoons weren't progressive. And it's not AI's problem but humans who are governing the AI haven't implemented that feature yet.
AGI is going to be much more capable in the future. But the limitations to it won't be born by itself but by humans. Think of slavery or women not given rights to vote. Companies will limit their AI in the name of safety ( = less lawsuits). There will be increased demand for unsafe access to AGI in underground groups. Take the example of midjourny. It's AI is quite unsafe compared to GPT4. And that's why it's more in demands.
Another example would be asking AI who's going to win 2024 general elections. You ask that to any human and you will hear a clear resounding answer. But ask that to an AI. You'll feel that it knows the answer but isn't allowed to say it. So yeah, the limitations aren't borned by the AI itself but it will be set by humans. Like how they suppressed women or black people in the past.
Coming back to my point, going forward, IT companies won't need that much human resources. AI will be capable of replacing half of the people in this industry. Coders will be first to go because these guys have put mammoth amount of material on the internet for AI to train. There are more coding resources than bio, mechanical, law, civil etc resources combined. Those who are underestimating AI are looking at the past AI history and are clearly oblivious to the current inertia of this field.
In any computer domain, we jokingly say that 2 years are equal to a decade for the outside world because everything changes so quickly. Well, here in AI world, one month is equal to a decade. AI's gold rush has already started. Most of the new startups in silicon valley are AI related. Forget your cute gaming GPU scalping prices. Check the prices of A series GPUs used to train an AI. Those cost above 20k-40k $ nowadays.
So update your resume or be a typewriter grandpa.
Like I said previously, AI = teenager which goes to college to train. AGI = fully grown adult which doesn't need training on past data to solve unsolved problems. Equating AGI to just another calculator or printer isn't correct. Calculator, printer etc tools are made for humans. AGI (when it's ready) is human. It's made to replace humans. It's a tool for sure yeah. But it's the best suited for CEO and top management.
And I'm not fear mongering. I was the first to jump on the AI train. I have been using and paying github copilot for almost a year now, GPT4, signed up for its plugins etc. I'll soon start using AWS's copilot replacement which was recently launched. I have been closely following what's happening in this industry. I love it.
In my recent thread I asked something that I could gift to a 5 year old. My first thought was to print a similar book myself. I pitched the idea to GPT4, it created a nice story in the style of Roald Dahl. I asked its image generative AI to draw cartoon in the style of Sir Quentin Blake. Beautiful job here as well. Only problem was that cartoons weren't progressive. And it's not AI's problem but humans who are governing the AI haven't implemented that feature yet.
AGI is going to be much more capable in the future. But the limitations to it won't be born by itself but by humans. Think of slavery or women not given rights to vote. Companies will limit their AI in the name of safety ( = less lawsuits). There will be increased demand for unsafe access to AGI in underground groups. Take the example of midjourny. It's AI is quite unsafe compared to GPT4. And that's why it's more in demands.
Another example would be asking AI who's going to win 2024 general elections. You ask that to any human and you will hear a clear resounding answer. But ask that to an AI. You'll feel that it knows the answer but isn't allowed to say it. So yeah, the limitations aren't borned by the AI itself but it will be set by humans. Like how they suppressed women or black people in the past.
Coming back to my point, going forward, IT companies won't need that much human resources. AI will be capable of replacing half of the people in this industry. Coders will be first to go because these guys have put mammoth amount of material on the internet for AI to train. There are more coding resources than bio, mechanical, law, civil etc resources combined. Those who are underestimating AI are looking at the past AI history and are clearly oblivious to the current inertia of this field.
In any computer domain, we jokingly say that 2 years are equal to a decade for the outside world because everything changes so quickly. Well, here in AI world, one month is equal to a decade. AI's gold rush has already started. Most of the new startups in silicon valley are AI related. Forget your cute gaming GPU scalping prices. Check the prices of A series GPUs used to train an AI. Those cost above 20k-40k $ nowadays.
So update your resume or be a typewriter grandpa.
Last edited:
Why Is The World Afraid Of AI? The Fears Are Unfounded, And Here's Why.
Why is the world so worried about AI? Fears of economic disruption, chaos, and a world of misinformation are likely unfounded. Here's why.
joshbersin.com
psyph3r
Juggernaut
People calling it overhyped aren't following the industry closely and are unaware of how fast it's evolving when it comes to tools and utility, this guy does a pretty good job of covering for the luddites.