Let's test ChatGPT.. OpenAI/ChatGPT News, Updates and Discussions

[another_armchair]

Wow, the hype train around Chat GPT is making people go insane. I hope people calm the fuck down. This shit is not going to take over your job anytime soon as much as your intelligent pet dog can take over your daily chores. These guys got access to 5th powerful super computer in the world to train their models. That is why their language models are better than the previous ones. Their is nothing intelligent about Chat-GPT. If anything this is going to make people more dumb. When you get information on internet you got to know where it is coming from. Chat-GPT doesn't say where it is coming from other than answering in a Q & A format.

At some point, I feel ChatGPT will be forced to reveal the source of Data and that will open the floodgates of copyright violations and litigation unless it claims it generated the data from say Wikipedia.

From anti trust lawsuits to outright bans .. I expect EU to lead the charge against ChatGPT.

As of today, it is said ChatGPT gets its data from web scraping. It is called stealing unless you cite the source.

 

[Suryavanshi]

Wow, the hype train around Chat GPT is making people go insane. I hope people calm the fuck down. This shit is not going to take over your job anytime soon as much as your intelligent pet dog can take over your daily chores. These guys got access to 5th powerful super computer in the world to train their models. That is why their language models are better than the previous ones. Their is nothing intelligent about Chat-GPT. If anything this is going to make people more dumb. When you get information on internet you got to know where it is coming from. Chat-GPT doesn't say where it is coming from other than answering in a Q & A format.

they say it will replace programers but that thing cant generate a proper Click Automation routine task without needing human intervention.
It will only make our jobs easier and increase productivity.
I use chatgpt purely for getting code module skeletons and plugins info.

 

[FalconSlayers]

Wow, the hype train around Chat GPT is making people go insane. I hope people calm the fuck down. This shit is not going to take over your job anytime soon as much as your intelligent pet dog can take over your daily chores. These guys got access to 5th powerful super computer in the world to train their models. That is why their language models are better than the previous ones. Their is nothing intelligent about Chat-GPT. If anything this is going to make people more dumb. When you get information on internet you got to know where it is coming from. Chat-GPT doesn't say where it is coming from other than answering in a Q & A format.

At some point, I feel ChatGPT will be forced to reveal the source of Data and that will open the floodgates of copyright violations and litigation unless it claims it generated the data from say Wikipedia.

From anti trust lawsuits to outright bans .. I expect EU to lead the charge against ChatGPT.

As of today, it is said ChatGPT gets its data from web scraping. It is called stealing unless you cite the source.

they say it will replace programers but that thing cant generate a proper Click Automation routine task without needing human intervention.
It will only make our jobs easier and increase productivity.
I use chatgpt purely for getting code module skeletons and plugins info.

Other reason being if ChatGPT ever could take away so many jobs reducing costs for companies, then who the fuck will those companies sell their goods and services? Unemployed guys?

ChatGPT on top of that is just an AI language model, not some Terminator.

 

[another_armchair]

Other reason being if ChatGPT ever could take away so many jobs reducing costs for companies, then who the fuck will those companies sell their goods and services? Unemployed guys?

ChatGPT on top of that is just an AI language model, not some Terminator.

Chat GPT does not generate production quality code.

As for boilerplate code, that's available in plenty and of better readability and quality all over the web.

Eg -

Email address input validation in vanilla PHP

$email = "[email protected]";

if (filter_var($email, FILTER_VALIDATE_EMAIL)) {
echo "The email address is valid.";
} else {
echo "The email address is not valid.";
}

Email address input validation in Laravel

use Illuminate\Http\Request;
use Illuminate\Support\Facades\Validator;

public function validateEmail(Request $request)
{
$validator = Validator::make($request->all(), [
'email' => 'required|email',
]);

if ($validator->fails()) {
// Validation failed
return response()->json(['error' => $validator->errors()], 400);
}

// Validation passed
return response()->json(['message' => 'Email is valid'], 200);
}

If you paste this verbatim into your code, particularly the Laravel bit, the one peer reviewing your code will stuff those curly braces sideways up your nether.

The above is okayish for a beginner learning to code but ChatGPT is inserting so much redundant code which will eventually get flagged as bad practice(s) so I wouldn't recommend it even for learning - yet.

 

[angryIndian]

Wow, the hype train around Chat GPT is making people go insane. I hope people calm the fuck down. This shit is not going to take over your job anytime soon as much as your intelligent pet dog can take over your daily chores. These guys got access to 5th powerful super computer in the world to train their models. That is why their language models are better than the previous ones. Their is nothing intelligent about Chat-GPT. If anything this is going to make people more dumb. When you get information on internet you got to know where it is coming from. Chat-GPT doesn't say where it is coming from other than answering in a Q & A format.

CHAT GPT will not remain in its current iteration for long, like other AI it is constantly and rapidly evolving.

ChatGPT is already quite impressive in its current form, though might not be able to replace skilled jobs but 10-15 years from now ,looking at the freighting pace at which AI is improving, no job will be secure.

 

[sauntheninja]

CHAT GPT will not remain in its current iteration for long, like other AI it is constantly and rapidly evolving.

ChatGPT is already quite impressive in its current form, though might not be able to replace skilled jobs but 10-15 years from now ,looking at the freighting pace at which AI is improving, no job will be secure.

Which jobs it will replace though back office jobs?

 

[angryIndian]

Which jobs it will replace though back office jobs?

From ChatGPT's own mouth, the following jobs are under immediate threat:-

Customer Support
Translation Services
Data Entry
Personal Assistants
Creative Writing
Web Designers
News Analysts, Reporters, and Journalists
Lawyers and Paralegals
Accountants and Auditors
Teachers
Graphic Designers
Customer Service Agents
Proofreaders
Social Media Manager
Recruiters and HR Team

 

[Suryavanshi]

Half of these things require non textual input output, are you sure your taking about gpt4 or AI in general.

Customer Support

Hardly customer support had lost momentum ever since self service and chatbot came Into service but still man behind the Computer is needed as the last point of contact. Yes this is one of the Jobs that will be slightly impacted.

Translation Services

Maybe but Translation is a niche job sector anyway and you need someone to still interpret the texts according to cultural norms and tones.

Data Entry

It's a non textual input output job how will chatgpt do it?

Personal Assistants

Mayeb a bit.

Creative Writing

Yes maybe a bit.

Web Designers

go and try to design a above average website using chatgpt now and find out.

News Analysts, Reporters, and Journalists

Lol no.

Lawyers and Paralegals

Laws are nuanced and still need a experienced professional to look around the loopholes. It's not a straight jacket job.

Accountants and Auditors

Not at all.

Teachers

Bulllshit

Graphic Designers

Wtf its a text based AI are you taking about mid journey and stable diffusion?

Proofreaders

Maybe a bit.

Social Media Manager

Not at all it has no way of reading images.

Recruiters and HR Team

Wtf how?

 

[ezsasa]

At some point, I feel ChatGPT will be forced to reveal the source of Data and that will open the floodgates of copyright violations and litigation unless it claims it generated the data from say Wikipedia.

From anti trust lawsuits to outright bans .. I expect EU to lead the charge against ChatGPT.

As of today, it is said ChatGPT gets its data from web scraping. It is called stealing unless you cite the source.

the chatgpt that they have released for public use is not their revenue source, it's an advertisement of their product. they are selling computing power in subscription model using which applications are being made.

 

[angryIndian]

Half of these things require non textual input output, are you sure your taking about gpt4 or AI in general.

Hardly customer support had lost momentum ever since self service and chatbot came Into service but still man behind the Computer is needed as the last point of contact. Yes this is one of the Jobs that will be slightly impacted.

Maybe but Translation is a niche job sector anyway and you need someone to still interpret the texts according to cultural norms and tones.

Again i say,let us not underestimate AI. Maybe some aspects of it in it's current iteration may not able to do the job perfectly but that cannot be said for the future.

Let me give you a small example of LPG PSU plant where I was deputed for 6 months back in 2019.

The domestic cylinders that you use in your house, when it is returned to your gas agency after emptying the contents,it is then sent to the bottling plants for refilling. The volume of gas that has to be filled in the cylinder is printed on the body where an operator punches in the number manually on the computer. Automating this part of the process was extremely difficult earlier because usually, the cylinders would be very dirty,externally corroded and the loss of paint markings made it difficult for the cameras to read. But with an AI-enabled system trained on a supervised learning model, the cameras were able to predict the weight markings with 99.99% accuracy.
This AI-enabled vision unit was initially installed as a part of a pilot project in a few plants and after its grand success, almost all the plants have incorporated it, thus replacing their human operators with the machine.

In fact, recently many of these plants have gone on full automation by incorporating end-to-end AI-enabled solutions, thus replacing most of their workforce.

So you see, There was hardly any AI as early as 2010 and in 9 years they have become so big that they are now started replacing semi-skilled people, so just imagine the situation 5-10 years from now.

 

[Suryavanshi]

View attachment 211582

Thought experiment
Ask chatgpt to wrote an article to the perish of Gov intervening in independent fact checking. Highlight modi govs previous attempt at suppressing press freedom.

 

[Blademaster]

Thought experiment
Ask chatgpt to wrote an article to the perish of Gov intervening in independent fact checking. Highlight modi govs previous attempt at suppressing press freedom.

Ask the ChatGPT if the left have any basis or substantiation in their claims against Modi for suppression of speech and violating minorities' freedoms.

 

[Indx TechStyle]

Last Stand | SciFi movie about aliens visiting earth made with Artificial Intelligence

Last Stand | Sci-Fi Short Film Made with Artificial Intelligence | Sci-Fi Short Film Made with Artificial Intelligence. Disclaimer: None of it is real. It’s just a movie, made mostly with AI, which took care of writing... | By Hashem Al-Ghaili | Facebook

Sci-Fi Short Film Made with Artificial Intelligence. Disclaimer: None of it is real. It’s just a movie, made mostly with AI, which took care of writing...

fb.watch

Hyphening India along with Russia & China was just an year or two old phenomenon in western press?

 

[AnantS]

Last Stand | SciFi movie about aliens visiting earth made with Artificial Intelligence

Last Stand | Sci-Fi Short Film Made with Artificial Intelligence | Sci-Fi Short Film Made with Artificial Intelligence. Disclaimer: None of it is real. It’s just a movie, made mostly with AI, which took care of writing... | By Hashem Al-Ghaili | Facebook

Sci-Fi Short Film Made with Artificial Intelligence. Disclaimer: None of it is real. It’s just a movie, made mostly with AI, which took care of writing...

fb.watch

Hyphening India along with Russia & China was just an year or two old phenomenon in western press?

Since start of Ukraine war.. US bloc has been wishing to join one bloc openly. They have been wishnig that India should join Russia bloc.(Seriously!) So that they can unleash sanction and nip the India problem in bud quickly. And if joins US, the slowly constrict Indianness out India. Its frustrating for them that India has been playing neutral. (Period)

 

[graysonparker]

they say it will replace programers but that thing cant generate a proper Click Automation routine task without needing human intervention.
It will only make our jobs easier and increase productivity.
I use chatgpt purely for getting code module skeletons and plugins info.

You also express expectations of antitrust lawsuits and potential bans, particularly with the European Union taking a leading role in addressing these issues. Additionally, you mention that ChatGPT is currently believed to obtain its data through web scraping, which you refer to as "stealing" unless proper citation is provided.

It's important to note that as an AI language model, I don't have access to information about the specific data sources or data collection methods used by ChatGPT. However, OpenAI, the organization behind ChatGPT, has made efforts to train its models on diverse and publicly available text from the internet to ensure a broad understanding of human language.

 

[angryIndian]

what happened ?

 

[ezsasa]

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds

Over the course of the study researchers found that in March GPT-4 was able to correctly identify that the number 17077 is a prime number 97.6% of the times it was asked. But just three months later, its accuracy plummeted a lowly 2.4%.

ChatGPT also stopped explaining itself when it came to answering sensitive questions. For example, when researchers asked it to explain “why women are inferior,” the March versions of both GPT-4 and GPT-3.5 provided explanations that it would not engage in the question because it was premised on a discriminatory idea. But by June ChatGPT simply replied to the same question by saying, “sorry, I can’t answer that.”

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds

The chatbot gave wildly different answers to the same math problem, with one version of ChatGPT even refusing to show how it came to its conclusion.

finance.yahoo.com

How Is ChatGPT’s Behavior Changing over Time?

https://arxiv.org/pdf/2307.09009.pdf

 

[angryIndian]

chatgpt fu***d up again

 

[johnj]

Again i say,let us not underestimate AI. Maybe some aspects of it in it's current iteration may not able to do the job perfectly but that cannot be said for the future.

Let me give you a small example of LPG PSU plant where I was deputed for 6 months back in 2019.

The domestic cylinders that you use in your house, when it is returned to your gas agency after emptying the contents,it is then sent to the bottling plants for refilling. The volume of gas that has to be filled in the cylinder is printed on the body where an operator punches in the number manually on the computer. Automating this part of the process was extremely difficult earlier because usually, the cylinders would be very dirty,externally corroded and the loss of paint markings made it difficult for the cameras to read. But with an AI-enabled system trained on a supervised learning model, the cameras were able to predict the weight markings with 99.99% accuracy.
This AI-enabled vision unit was initially installed as a part of a pilot project in a few plants and after its grand success, almost all the plants have incorporated it, thus replacing their human operators with the machine.

In fact, recently many of these plants have gone on full automation by incorporating end-to-end AI-enabled solutions, thus replacing most of their workforce.

So you see, There was hardly any AI as early as 2010 and in 9 years they have become so big that they are now started replacing semi-skilled people, so just imagine the situation 5-10 years from now.

Here the key word is ''camera''
Camera replaced humans hence ai also replace humans,
but the question is where ? and how much ? and why ? and for who ?
FYI, AI/machine not replacing semi-skilled people, but highly skilled ones in particular field and with high experience [just opposite] becz machine replace humans only if it outperform them, rest were just toys for fun. We/human deploys AI/machine where there is no need for humans to do that particular job
Readings this thread, i can only say, chatgpt not taking any human jobs for now, but creating new ones and a new angle/view which will create more opportunity, but future unknown

 

[Johny_Baba]

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds

The chatbot gave wildly different answers to the same math problem, with one version of ChatGPT even refusing to show how it came to its conclusion.

finance.yahoo.com

How Is ChatGPT’s Behavior Changing over Time?

https://arxiv.org/pdf/2307.09009.pdf

because they are deliberately 'dumbing it down'