ChatGPT - More revolutionary than Google

[Krash]

Originally Posted by Mifune

How many of you realize that Open AI is an Elon Musk domain?

Microsoft is now the big money behind OpenAI. $20 billion investment plus free use of Microsoft Cloud /Azure.

Elon Musk did provide $1 billion worth of initial funding, but is no longer affiliated with OpenAI. He left the board of OpenAI back in 2018.


Sent from my iPhone using Tapatalk

[Window Shopper]

Google invests $300M in ChatGPT rival:
https://www.linkedin.com/news/story/...rival-6154898/

[brandrea]

Originally Posted by Krash

…
It's amazing. High school and college kids can use it to generate essays on virtually anything...

Interesting indeed

I wonder about this though, as I understand it colleges and universities use some sort of algorithm programs to detect plagiarism, so you’d have to think too many essays generated this way would not work out so well

It is a very cool technology for sure though. Scary to me in a lot of ways.

We toured the library at our daughter’s university last month, it it didn’t resemble anything like I was familiar with…. Very few books, and no microfiche

[DLRIDES]

Originally Posted by Mifune

How many of you realize that Open AI is an Elon Musk domain?

…………… and ?

…………… please expand.

[Krash]

Originally Posted by DLRIDES

…………… and ?

…………… please expand.

He's not. Elon Musk resigned from OpenAI's board back in 2018. But I'm thinking he regrets doing so.

OpenAI just seems like the type of thing he'd like to control.

[brandrea]

Originally Posted by Phrank

A Comp Sci student here in Toronto just came up with another app that can identify anything written using "ChatGPT", teachers have been downloading it in pretty big number the article says.

So aside from the database they have to prevent plagiarism in use now, I'm sure it won't be long until most papers have to also pass this test as well....

Sorry I didn’t see your post earlier, but this is my understanding as well.

[Krash]

ChatGPT is really good at giving responses that will trigger people on TRF. LOL!




Sent from my iPhone using Tapatalk

[mountainjogger]

Originally Posted by Krash

Compared to Google, the striking thing about ChatGPT is how it gives a single concise response to a query. Google on the other hand will give hundreds, possibly thousands of responses (plus ads), and then you have to scroll through them to get the information you want.

A good analogy is comparing both to a library. With ChatGPT, you only need to read one book to get what you want. With Google, you might have to look at dozens and dozens of books to get the information you want.


Sent from my iPhone using Tapatalk

What you are really getting is the book the AI has decided you want based on volume data of other from other people. It may or may not be the book you want.

This is fine for some things if you are willing to accept cliff notes through the AI's filter - amalgamated from what most others have said or written, or searched for. It is also great for people who lack the ability or patience to scan large amounts of information quickly, or the desire to come up with an original thought. Which sadly, is an ever increasing percentage of the population.

But if you are looking for originality or innovation, it is lacking. And if you are looking for a needle in a haystack, it is lacking.

For example, in my profession, AI assisted brief writing for lawyers has been available for a while. You can also increase the level of AI involvement in your legal research.

AI assisted briefs lack the nuance, style, passion and impact of a truly good writer. They also lack the context/judgment calls of taking into consideration the audience (who is the judge or panel?) and the politics of the moment (do you really want to make that argument given what was on the news last night?

But where AI assistance really breaks down is when you have novel issues or facts of first impression. And unless you are doing volume cookie cutter work, these issues crop up pretty regularly, if you are diligent enough to look for them.

We are a long way from an AI program that can go back to basics, weed through various fundamental principals and then make judgment calls as to which ones and how to apply them to a new situation or argue by analogy, with a mind on the audience, and the context.

[Krash]

Originally Posted by mountainjogger

What you are really getting is the book the AI has decided you want based on volume data of other from other people. It may or may not be the book you want.

This is fine for some things if you are willing to accept cliff notes through the AI's filter - amalgamated from what most others have said or written, or searched for. It is also great for people who lack the ability or patience to scan large amounts of information quickly, or the desire to come up with an original thought. Which sadly, is an ever increasing percentage of the population.

But if you are looking for originality or innovation, it is lacking. And if you are looking for a needle in a haystack, it is lacking.

For example, in my profession, AI assisted brief writing for lawyers has been available for a while. You can also increase the level of AI involvement in your legal research.

AI assisted briefs lack the nuance, style, passion and impact of a truly good writer. They also lack the context/judgment calls of taking into consideration the audience (who is the judge or panel?) and the politics of the moment (do you really want to make that argument given what was on the news last night?

But where AI assistance really breaks down is when you have novel issues or facts of first impression. And unless you are doing volume cookie cutter work, these issues crop up pretty regularly, if you are diligent enough to look for them.

We are a long way from an AI program that can go back to basics, weed through various fundamental principals and then make judgment calls as to which ones and how to apply them to a new situation or argue by analogy, with a mind on the audience, and the context.

Yes, when it comes to highly esoteric subject matter, the responses aren't exactly PHD level in most cases. Not even close.

I was messing around with it. It's pretty high level. It's good at giving an introductory (101) level repsonse, but the more complex the subject matter, the more disapointed you'll be with the response.

Also, we talked above about Google not exactly being a reliable source either.

[brandrea]

Originally Posted by Krash

ChatGPT is really good at giving responses that will trigger people on TRF. LOL! …

What does it say about whether it’s safe to wear a Rolex in LA?

[Krash]

Originally Posted by brandrea

What does it say about whether it’s safe to wear a Rolex in LA?

Sent from my iPhone using Tapatalk

[brandrea]

Originally Posted by Krash

Sent from my iPhone using Tapatalk

I think this AI is a member here

Sent from my SM-G960W using Tapatalk

[TswaneNguni]

Originally Posted by austinp

Unfortunately it looks to me like another bit of tech that prevents people from thinking for themselves.

Fewer and fewer can .

[Drake20]

It’s my first time hearing about ChatGpt. I think it’s better than google anyways


Sent from my iPhone using Tapatalk

[BLACKHORSE 6]

Originally Posted by INC

No. With ChatGPT, you won't get the book what you want but what they want you to get. This is the path which leads to the complete demolition of the human beeing as this techology eliminates the possibility to build YOUR OWN opinion with the help of knowing different aspects and arguments.

There is nothing new in this except the public domain. Think about to Cambridge Analytica and Brexit, or even the US presidential elections. This is not a brave new world, but the opening of the highway to hell.

Welcome to the real world Neo...

This.

“Someone” is in control of the algorithm that dictates the answers that ChatGPT spits out. That worries the crap out of me. It’s the same with Google and other tech apps, but people are going to take the answers of ChatGPT and turn them into gospel.

[Krash]

Originally Posted by BLACKHORSE 6

This.

“Someone” is in control of the algorithm that dictates the answers that ChatGPT spits out. That worries the crap out of me. It’s the same with Google and other tech apps, but people are going to take the answers of ChatGPT and turn them into gospel.

Satya Nadella is pushing a lot of the buttons, so we should be very scared about this.


Sent from my iPhone using Tapatalk

[Kowaco]

I just started using it, in fact it wrote this post for me.


Sent from my iPad using Tapatalk

[RichardBartlett]

Originally Posted by Maleg

Students using this to write their papers prepares them for what? They don’t have to learn how to research, organize thoughts in a logical way, or express thoughts in a clear and concise way. They will inevitably be supervised by this tool. Which, will replace all the MBAs and junior managers at every major corporation.

The next two decades are going to see massive changes in the job market. As robots did for manufacturing blue collar jobs so this and others will wipe out large sections of white collar jobs. Those people who thought they were indespensible will find out how untrue that is. Accountants, research scientists, lawyers all those who thought brains were worth more than brawn will come to a new reality.

The world will need to find a way to pay most people not to work and a few people to work.. Going to be very bumpy imho

[Krash]

Originally Posted by Window Shopper

Microsoft will let companies create their own custom versions of ChatGPT.
Source: https://www.cnbc.com/2023/02/07/micr...ze-source.html

That makes sense. Microsoft has invested billions and provides OpenAI / ChatGPT with free use of servers and infrastructure. But even after all that, they still don’t technically own it.


Sent from my iPhone using Tapatalk

[Blansky]

Seems kind of unfair that spreaders of conspiracy theories and mis-information will now be put out of a job by this new technology which can write it better with less spelling errors.

I guess they can always go back to writing advertising copy or playing video games in their mom's basement.

[Blansky]

A couple of articles....

https://newatlas.com/technology/chat...e-replacement/

https://www.nytimes.com/2023/02/08/t...a98d9ddd0088d2

[Star Ferry]

Originally Posted by Blansky

Seems kind of unfair that spreaders of conspiracy theories and mis-information will now be put out of a job by this new technology which can write it better with less spelling errors.

I guess they can always go back to writing advertising copy or playing video games in their mom's basement.

ChatGPT mostly produces copy that self-appointed misinformation experts would agree with, but yes, AI technology could be adapted to produce copy that they disagree with.

High-level view, I wish folks could take a broader view on what constitutes misinformation or a conspiracy theory.

I know people who work at (or have worked at) all of the media sources you think are credible (NYT, The Atlantic, CNN, MSNBC) -- they're all millennials from affluent families -- and only a few of them trust these sources as much as you do. Most believe that some of their employer's content is legitimately informative, while some of it is BS. Others believe it's mostly nonsense, but go along with it either because they like having a cool job, or because they need to have some kind of job lest they be fully exposed as a trust fund baby. I also know plenty of people at sources you don't trust (like Fox News) -- they are similar people with similar views about their employer.

[DocHorton]

Originally Posted by Blansky

Seems kind of unfair that spreaders of conspiracy theories and mis-information will now be put out of a job by this new technology which can write it better with less spelling errors.

I guess they can always go back to writing advertising copy or playing video games in their mom's basement.

True. Google, FB, and the MSM are going to have a hard time with it.

[Window Shopper]

The tech is very good, but it needs to be updated. It says:
"May occasionally generate incorrect information"
"May occasionally produce harmful instructions or biased content"
"Limited knowledge of world and events after 2021"

Yesterday I asked it for "the best hotels in the world". It outdatingly said one of them was 'Mandarin Oriental, Las Vegas', but then I Googled to find it was rebranded to 'Waldorf Astoria Las Vegas'. Today I asked "what brands do Chanel own" (after watching the Chanel movie Coco Before Chanel), and it incorrectly said 'Fendi and Boucheron', but they're actually owned by Kering.

[77T]

Why do people love ChatGPT?

She told me this:
(“She who must be obeyed”)

“People love ChatGPT because it is a highly advanced artificial intelligence language model that can understand and generate human-like text. It is trained on a massive amount of diverse data, making it knowledgeable about a wide range of topics, from science and history to pop culture and current events. As a result, ChatGPT is able to answer questions, provide explanations, and even engage in conversations in a way that feels natural and human-like.

Many people find ChatGPT to be a valuable tool for answering questions, helping with decision making, and even for entertainment purposes. The ability to have a conversation with a highly knowledgeable AI is a unique and exciting experience that many people enjoy. Additionally, because ChatGPT is a machine learning model, it continues to improve and evolve as it is used and exposed to more data, making it an even more valuable resource over time.”


Sent from my iPhone using Tapatalk Pro

[Star Ferry]

Originally Posted by Window Shopper

The tech is very good, but it needs to be updated. It says:
"May occasionally generate incorrect information"
"May occasionally produce harmful instructions or biased content"
"Limited knowledge of world and events after 2021"

Yesterday I asked it for "the best hotels in the world". It outdatingly said one of them was 'Mandarin Oriental, Las Vegas', but then I Googled to find it was rebranded to 'Waldorf Astoria Las Vegas'. Today I asked "what brands do Chanel own" (after watching the Chanel movie Coco Before Chanel), and it incorrectly said 'Fendi and Boucheron', but they're actually owned by Kering.

It’s mainly good at generating word salad. I asked it to write me an essay about resilience and perseverance, as well as an essay about the green light in The Great Gatsby, and it delivered more than most middle schoolers would need for a B+ or an A-.

[huncho]

people will forget about it soon. it doesn't provide much value in its current state and is basically just a viral meme on social media, eventually they'll find a new flavor of the month and move on. i can see how kids flocked to it to have papers written but they won't get away with it for much longer, if they still are (i assume they're not)

[Krash]

Originally Posted by huncho

people will forget about it soon. it doesn't provide much value in its current state and is basically just a viral meme on social media, eventually they'll find a new flavor of the month and move on. i can see how kids flocked to it to have papers written but they won't get away with it for much longer, if they still are (i assume they're not)

I really disagree. ChatGPT just came out and there is BIG MONEY behind it. It only took 1 week to have a million users. That’s way faster than FB, Twitter, Amazon, Google and Netflix. It’s going to get better and better and better.

It’s also a MASSIVE threat to Google. Many people will prefer ChatGPT for simple search requests. Plus, people would love to get away from Google anyway, but there just weren’t any other options until now. With that said, I’m concerned about the implications of a tool like this.

Quick sample of search capabilities:




Sent from my iPhone using Tapatalk

[huncho]

Originally Posted by Krash

I really disagree. ChatGPT just came out and there is BIG MONEY behind it. It only took 1 week to have a million users. That’s way faster than FB, Twitter, Amazon, Google and Netflix. It’s going to get better and better and better.

It’s also a MASSIVE threat to Google. Many people will prefer ChatGPT for simple search requests. Plus, people would love to get away from Google anyway, but there just weren’t any other options until now. With that said, I’m concerned about the implications of a tool like this.

Quick sample.




Sent from my iPhone using Tapatalk

it could become something big but right now it just seems like everyone's playing around with it because it's viral. also google isn't gonna just pull a yahoo and fade away, they have more than enough data and tech to bring out something much better than this that actually aggregates searches combined with personalized data and machine learning dialogue

there's also tons of examples of it showing bias because of the way it's designed, at least with google searches you have lots of real opinions and forums to go off of. from what i understand ChatGPT isn't actually connected to the internet, it basically is just like a robot that learns through dialogue and questions/answers, which is impressive given the answers it's able to give but it's not really a replacement for google in my eyes. what makes google so good is it indexes the entire internet and then also customizes things because it sees everything you do. i see this having more of a threat to certain jobs that involve writing but less on google and searching personally. it will be interesting to see what kind of tech this pushes to be created but i feel like google has such a huge lead in this space that it's theirs to lose

[Krash]

Originally Posted by huncho

it could become something big but right now it just seems like everyone's playing around with it because it's viral. also google isn't gonna just pull a yahoo and fade away, they have more than enough data and tech to bring out something much better than this that actually aggregates searches combined with personalized data and machine learning dialogue

there's also tons of examples of it showing bias because of the way it's designed, at least with google searches you have lots of real opinions and forums to go off of. from what i understand ChatGPT isn't actually connected to the internet, it basically is just like a robot that learns through dialogue and questions/answers, which is impressive given the answers it's able to give but it's not really a replacement for google in my eyes. what makes google so good is it indexes the entire internet and then also customizes things because it sees everything you do. i see this having more of a threat to certain jobs that involve writing but less on google and searching personally but it will be interesting to see what kind of tech this creates but i feel like google has such a huge lead in this space that it's theirs to lose

We’re not going to solve this tonight, but I view this as one of biggest disruptors I’ve ever seen.

One thing for certain, when the IPO comes out, I’m putting my money where my mouth is. I’m buying as many shares as I can.

Also, I’m pretty sure it’s connected to the internet. I’ve seen prototypes where it lists the websites used to produce the results. So that’s coming down the road.


Sent from my iPhone using Tapatalk