Professions that are screwed pretty soon because of AI

[TeTe]

When it's sufficiently good it can solve problems with metaphorical leaps, so taking a solution from one problem and applying it to another, which is exactly what our innovation is.

But, even if you think that is some kind of voodoo that humans will keep, 99% of human work is just repetition.

Nothing is ever new, it's always an amalgamation of befores information with the odd lateral movement. I'm surprised people think that's beyond automation.

Cool. Can't wait till we're all out of work... And the companies that only use AI go bankrupt because they have no consumers to buy their services or products.

It's a great reason to cheerlead for a dystopian society.

 

[cincymma79]

So what you're saying is, humans create the AI and program it to be just as shitty as humans. I mean, I guess it's maybe bring more efficient at being shitty.

 

[djacobox372]

It is an incredibly powerful tool. But any tool is just a tool unless it starts making decisions on its own.
And I don't mean a decision like what number to set as a credit score for a bank's client or how to draw a magical forest based on a prompt.
I mean a kind of decision when you prompt an AI tool to paint a magical forest and instead it makes a melody. Or decides to do fuckall because it doesn't want to paint a forest. Those are two very different levels of thinking and decision making. The former is an advanced tool and nothing more, the latter is intelligence in the full meaning if the word. Modern AI cannot do the latter.

So intelligence is not doing what you're told?

If I ask an artist to paint me a forest and he delivers a fucking melody, or does "fuckall" that artist is fucking fired. LOL

Not doing what your asked isn't intelligence, it's free-will.

 

[avenue94]

I can't comment on "what you see". That is not me. Sounds like you think AI is a scam though. Is that right?

If you mean presenting LLM-based developments as actual artificial intelligence, yes, that's just marketing. If you mean that the entire current kit and caboodle is a scam, no, lots of potential there. However, most of the hype is from bandwagoners who don't know shit about the actual products or how they are made.

Data is not static. The AI themselves are generating exponentially more data than we had prior. Self driving cars. Wearable devices. Fitness trackers. Your cell phone. FinTech, Cloud....

The data being generated, tracked and aggregated currently dwarfs the entirety of what we have historically.

Are you unaware of this?

Data cannot be created overnight. Current models are based off effectively everything humans have created to date. We would have to wait a long time to get a data set of similar quantity and quality.

Hint: self driving data isn't particularly useful for Chat GPT, it's only useful for self driving and related fields.

You are aware that 5.0 is way behind schedule and why, right?

Wrong. They are not simply recognizing patterns as that would mean the pattern already exists. They are creating patterns such as new languages and other things that do not resemble anything mankind has created before. You are simply wrong.

What has AI created that doesn't resemble anything humans have made?

 

[avenue94]

you pay for the license

Possible but it's unlikely to cover the costs for everyone in the pipeline. Consumers don't pay anywhere near market rate costs, and enterprise pricing is also sub-market rate even when it ramps up steeply after you get pass a certain amount of tokens. It's not entirely clear businesses and customers will actually pay market rate for current tools. That's also setting aside the sheer amount of AI washing that goes on in the industry (ie AI=All Indians).

they can make you sign shit that they get % of your gains, easiest method would be you can sell stuff made with their tool only on their own marketplace, so they will automatically be able both to check sales and hold their slice of money

This will never fly, too expensive for customers. There's a reasons enterprise licenses rarely work like this for big money projects.

Put it this way: How many big AI players are in the black right now, aside from the silicon level companies?

 

[avenue94]

So intelligence is not doing what you're told?

If I ask an artist to paint me a forest and he delivers a fucking melody, or does "fuckall" that artist is fucking fired. LOL

Not doing what your asked isn't intelligence, it's free-will.

It's not frere will in the case of AI, it's AI being too dumb. It's like if I told a 5 year old to paint a rainbow and they've never seen one and paint a rock.

 

[TribalDrumz]

Jeez @TeTe you went HAM in here

Every time I hear AI reminds me of my Allen Iverson jersey that got stolen

 

[Phlog]

Cool. Can't wait till we're all out of work... And the companies that only use AI go bankrupt because they have no consumers to buy their services or products.

It's a great reason to cheerlead for a dystopian society.

Yes, we're going to have to be quite smart to navigate it. I hope that the Christmas spirit prevails and jesus sorts it all out.

 

[ChosenOne]

If you mean presenting LLM-based developments as actual artificial intelligence, yes, that's just marketing. If you mean that the entire current kit and caboodle is a scam, no, lots of potential there. However, most of the hype is from bandwagoners who don't know shit about the actual products or how they are made.

Data cannot be created overnight. Current models are based off effectively everything humans have created to date. We would have to wait a long time to get a data set of similar quantity and quality.

Hint: self driving data isn't particularly useful for Chat GPT, it's only useful for self driving and related fields.

You are aware that 5.0 is way behind schedule and why, right?

Data IS created over night, each and every night.

Hint not needed. NO. You continue to show you do not know what you are talking about. Self driving car data does not have a sole use only for self driving cars and related fields, as you claim. No data really works that way. Retail will end up pulling data from it. As will banking. And so on and so on. Is it most relevant to the self driving car industry, sure. Just like banking data is most relevant to that industry.

And again you need to just stop speaking on this topic as you are massively ignorant about it. You seem to not know that most of the data today was created in the last 2 years and that is scaling WAY up now.

Top Data Created Stats (Editor's Choice)​

• Approximately 402.74 million terabytes of data are created each day
• Around 147 zettabytes of data will be generated this year
• 181 zettabytes of data will be generated in 2025

Data Creation Growth Projections​

The amount of data generated annually has grown year-over-year since 2010.

In fact, it is estimated that 90% of the world's data was generated in the last two years alone.

In the space of 13 years, this figure has increased by an estimated 74x from just 2 zettabytes in 2010.

The 120 zettabytes generated in 2023 are expected to increase by over 150% in 2025, hitting 181 zettabytes.

What has AI created that doesn't resemble anything humans have made?

That is not a meaningful question. If AI cures diseases like cancer, when mankind could not, that still resembles things humans have done, so what is the point of your question? Why does AI have to create something that does not even resemble anything humans have done?

 

[avenue94]

Data IS created over night, each and every night.

Hint not needed. NO. You continue to show you do not know what you are talking about. Self driving car data does not have a sole use only for self driving cars and related fields, as you claim. No data really works that way. Retail will end up pulling data from it. As will banking. And so on and so on. Is it most relevant to the self driving car industry, sure. Just like banking data is most relevant to that industry.

And again you need to just stop speaking on this topic as you are massively ignorant about it. You seem to not know that most of the data today was created in the last 2 years and that is scaling WAY up now.

What other uses does self driving data have industry wide for training?

And most data that OpenAI used is much older, even if it was scraped recently. You also are mistakenly under the impression that all the data in datacenters (per your link) is able to used. It is not. There is a reason most companies don't put proprietary data into public models.

Let me ask you this. You claim to be an investor right? What's the core AI product you invested in? Because I get the feeling you are very far from the actual product and don't actually know how the sausage is made. I'm no expert but I work in the data industry and see a lot more of this than your average person.

 

[Cole train]

If everyone loses their jobs, how will companys earn profit since no one has money to buy their products?

 

[Noveltymortar]

If everyone loses their jobs, how will companys earn profit since no one has money to buy their products?

Universal basic income. We will all be like welfare pets for the controlling classes in the future.

 

[Cole train]

Universal basic income. We will all be like pets for the controlling classes in the future.

So you are saying i can afford rent food hobbies without working?

Fucking sweet , AI destroy employment market please

 

[Noveltymortar]

So you are saying i can afford rent food hobbies without working?

Fucking sweet , AI destroy employment market please

Yes sir. You are about to hit the jackpot. You will also get a monthly government vodka allowance.

 

[chardog]

Why do you believe this?

Did machine learning exist in any meaningful way a few years ago? No... it is something new.

Why can't AI learn to create original works given the right set of parameters? What would prevent that?

Yes it has existed, and the parameter list would be too large to list. Humans can read between the lines a whole lot better than computers.

Having AI create simple application functions in itself can be a chore, and it doesn’t mean the function itself should exist. AI is great for spitting out one off references that i would have to research since google is completely broken.

 

[ChosenOne]

What other uses does self driving data have industry wide for training?

And most data that OpenAI used is much older, even if it was scraped recently. You also are mistakenly under the impression that all the data in datacenters (per your link) is able to used. It is not. There is a reason most companies don't put proprietary data into public models.

Let me ask you this. You claim to be an investor right? What's the core AI product you invested in? Because I get the feeling you are very far from the actual product and don't actually know how the sausage is made. I'm no expert but I work in the data industry and see a lot more of this than your average person.

It is clear you are no expert. Your questions in this area ar child like. Data is NOT locked to the segment it is created in. It gets aggregated and used in ways you will never see. If you think Bank data is only used by banks and retail data only by retailers, you really are foolish.

I am not an expert either but i do associate regularly with people who are. i was very involved with an AI incubator program at one of the top Universities as an entrepreneur in residence, in this area, when AI was really emerging. I can text right now a person considered in the Top5 fathers of Ai and he will text me back almost immediately.

And again you are simply wrong in thinking these large AI data sets are mostly older legacy data and they are not also drawing from the much more abundant data that has ramped up in the last 5 years, such as from eBanking, Wearable device data, Fitness data, Retail data, auto data, Smart home data, smart phone data, Social Media data, et, etc, etc,

ALL OF THIS DATA is being aggregated and compiled and used in these huge database sets and we have entered a new Data age that is vastly eclipsing what we had prior.

And yet here you are arguing the data of the future will not elipse the historical data despite the FACT 90% of todays data was developed in the last two years. And data creation is GROWING year over ear at a fast rate and here you are still saying 'nope the historical data will dominate for the foreseeable future'.

Your position is laughably ignorant. Not worthy of serious discussion.

 

[chardog]

I can't comment on "what you see". That is not me. Sounds like you think AI is a scam though. Is that right?

Data is not static. The AI themselves are generating exponentially more data than we had prior. Self driving cars. Wearable devices. Fitness trackers. Your cell phone. FinTech, Cloud....

The data being generated, tracked and aggregated currently dwarfs the entirety of what we have historically.

Are you unaware of this?


Wrong. They are not simply recognizing patterns as that would mean the pattern already exists. They are creating patterns such as new languages and other things that do not resemble anything mankind has created before. You are simply wrong.

You are a salesman for AI, and that surely is profitable because of the hype, however the use case is limited and the end goal is enhancing worker experience rather than replacing them because the tools are underwhelming.

For professional fields, the buck has to stop with someone, blaming AI failures only works if you’re a nefarious health care company.

 

[ChosenOne]

How Big Data in Autonomous Vehicles Defines the Future​

...

What are the use cases of big data autonomous vehicles?​

Think that only smart cars themselves use the data they gather? Think again. The true potential of this data is much greater:

• An automaker can remotely see a problem with your car and immediately inform you through your vehicle.
• With vehicle data, managing a fleet of hundreds or thousands of connected cars becomes an efficient and optimized process.
• Big data is a rich source of behavioral insights, from the gas stations drivers choose to the music they listen to. This information can be used in marketing, sales, and customer service.
• City planning and engineering will become more accurate with big data from connected vehicles: more efficient road planning, early warning systems on dangerous spots, and safer pedestrian walkways are just a few possible outcomes.
• Big data in connected cars can help to develop custom insurance plans for drivers based on their performance.

 

[ChosenOne]

You are a salesman for AI,

If that means i clear up misinformation then yes.

and that surely is profitable because of the hype,

the hype is definitely profitable as is the actual use cases.

however the use case is limited

If you believe that you are fool. We are only at the beginnings of what it can do comparable to the first cell as it evolved to the smart phone, Use cases in surgery already are allowing surgeries and detections that could not be done prior, as just one instance.

and the end goal is enhancing worker experience

that is part of the goal.

rather than replacing them because the tools are underwhelming.

ya you are fool speaking so definitively and factually on something you know so little about and are wrong about. Whether it be in some surgeries or computer programming tasks for two areas of many i could name it is already replacing jobs.

For professional fields, the buck has to stop with someone, blaming AI failures only works if you’re a nefarious health care company.

Ok. NOt sure what point you think you are making but i basically agree with it.

 

[TeTe]

You are a salesman for AI, and that surely is profitable because of the hype, however the use case is limited and the end goal is enhancing worker experience rather than replacing them because the tools are underwhelming.

You didn't know Sam Altman posted in sherdog?