ChatGPT AI is a huge breakthrough, a Google killer

[wade1mil]

I'm not sure if it was mentioned in this thread, but I read somewhere a week or two ago that Google already has something similar to ChatGPT and they're four years ahead of them. I tried finding the article, but was unsuccessful. I scanned the article at the time and it mentioned that Google had a concern that has stopped them from releasing it, but I don't recall what their concern was.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[mikecarlooch]

I'm not sure if it was mentioned in this thread, but I read somewhere a week or two ago that Google already has something similar to ChatGPT and they're four years ahead of them. I tried finding the article, but was unsuccessful. I scanned the article at the time and it mentioned that Google had a concern that has stopped them from releasing it, but I don't recall what their concern was.

4 years ago google showcased something that is still not public which was an AI that sounds exactly like a human. I believe they had the AI book an appointment for a haircut and the person on the other end had no idea they were talking to a bot.

insane that this technology already exists but hasn't been released to public.

Will we be able to automate sales calls with AI soon..?

 

[DCG]

My bot vs your bot?

Gonna reiterate that this is a great danger to Google and standard search engines. Even saw a few articles that said as much.

I don't think I've needed to use a search engine in days, I just now use this, and the answers are much easier to find. The response here is a great example as it told me exactly what it was.

View attachment 46505

If you use could AI to produce more instead of just using it to replace you, then you could make bank with it

 

[Skroob]

4 years ago google showcased something that is still not public which was an AI that sounds exactly like a human. I believe they had the AI book an appointment for a haircut and the person on the other end had no idea they were talking to a bot.

insane that this technology already exists but hasn't been released to public.

Will we be able to automate sales calls with AI soon..?

Mobile technology is my industry, so I remember this demo well. It was, frankly, absolutely stunning when it happened. The tech is called Google Duplex, and Google bills it as an AI assistant for tedious multistep tasks. There is a web version, which is apparently deprecated as of this month, but the Android phone version is alive and well. At least, that's what I've been able to determine from searches. It looks like "Hold for Me" and "Direct My Call" are the two major things that Duplex does on Pixel devices, but I haven't heard much about its ability to make haircut appointments for you since I/O 2018.

Related, there was a decent amount of speculation at the time that the demo was either entirely faked or, at minimum, scripted or edited. (See this contemporary article on Axios: What Google isn't telling us about its AI demo) I don't know if those questions ever got answered, but the fact that people aren't having Voice of Mountain View do all their tedious phone calls for them feels like some kind of an answer.

(For fun, and to stoke the fires of the iPhone/Android holy war, contrast with Steve Jobs introducing the original iPhone, where he calls a local Starbucks and orders 4000 lattes to go. This was apparently completely unscripted , to the point where they did not tell the Starbucks that it was going to happen. The first live public phone call ever made on an iPhone was a prank call.)

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[akshat.jha21]

Been spending some time on ChatGPT, holy shit this thing is insane, a total Google killer.

It in effect, is like the Star Trek "computer" for those of you familiar with Star Trek.

This thing already likely saved me 3 hours of my time in a matter of 10 seconds.

I haven't been this impressed with a new technology in 15 years, the last time being the Apple iPhone replacing a flip-phone.

It can spit out computer code for you in seconds, stuff that would take 10 different Google searches amount 100 different results filled with SEO manipulated filler and garbage.

If you're a writer, this thing will help you describe scenes in foreign countries, write dialogue in different types of personalities, and conquer blocks in a matter of seconds.

The utility in this application is beyond belief.

The old way: 1) Search Google, 2) Sift through mountains of "SEO'd" pages filled with filler 3) Spend 20 minutes finding the right passage on the right page.

Time: 15m

The new way: 1) Ask ChatGPT

Time: 15s

Here is what ChatGPT responded with how to make money.

I couldn't have answered better myself.

I have been using ChatGPT a lot lately. IT IS AMAZING. I'll probably stop using Google search and start using this full-time.

The way the AI scraps through large amounts of data are insane. It takes some time to type out the answer, but the quality is superior to any search engine out there.

Whether you're a student writing your assignments, a programmer finding some code or an entrepreneur finding some good advice, ChatGPT got you covered.

 

[Andreas Thiel]

I have been using ChatGPT a lot lately. IT IS AMAZING. I'll probably stop using Google search and start using this full-time.

The way the AI scraps through large amounts of data are insane. It takes some time to type out the answer, but the quality is superior to any search engine out there.

Whether you're a student writing your assignments, a programmer finding some code or an entrepreneur finding some good advice, ChatGPT got you covered.

People seem to to end up on a different version of ChatGPT ... I see outtakes and run into walls all the time. Outtake tweets on Twitter are also on the rise.

 

[Xeon]

I'm not sure if it was mentioned in this thread, but I read somewhere a week or two ago that Google already has something similar to ChatGPT and they're four years ahead of them. I tried finding the article, but was unsuccessful. I scanned the article at the time and it mentioned that Google had a concern that has stopped them from releasing it, but I don't recall what their concern was.

If they really had that, then they won't be in panic mode now: Google's management has reportedly issued a 'code red' amid the rising popularity of the ChatGPT AI

Companies make bold claims and smoke/mirrors to drive up their stock value time to time.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[wade1mil]

If they really had that, then they won't be in panic mode now: Google's management has reportedly issued a 'code red' amid the rising popularity of the ChatGPT AI

Companies make bold claims and smoke/mirrors to drive up their stock value time to time.

True. The media also knowingly makes false claims to politically and financially assassinate companies and people. Time will tell.

 

[Andy Black]

I have been using ChatGPT a lot lately. IT IS AMAZING. I'll probably stop using Google search and start using this full-time.

The way the AI scraps through large amounts of data are insane. It takes some time to type out the answer, but the quality is superior to any search engine out there.

Whether you're a student writing your assignments, a programmer finding some code or an entrepreneur finding some good advice, ChatGPT got you covered.

It's a great tool. I still prefer finding authoritative sites and reading their content though. I especially miss an actual opinion and people pinning their flag to a mast.

Note that if everyone uses ChatGPT instead of Google it's reducing visits to the websites people found via Google. So those sites are getting scraped but not attributed to or rewarded with a visit? What's their incentive to post any new content if AI chat will scrape it and not attibute or link back to it? It will be interesting to see how this pans out.

 

[akshat.jha21]

It's a great tool. I still prefer finding authoritative sites and reading their content though. I especially miss an actual opinion and people pinning their flag to a mast.

Note that if everyone uses ChatGPT instead of Google it's reducing visits to the websites people found via Google. So those sites are getting scraped but not attributed to or rewarded with a visit? What's their incentive to post any new content if AI chat will scrape it and not attibute or link back to it? It will be interesting to see how this pans out.

Yes, you are right. I would say ChatGPT is a good choice if you are looking to save time. But if good quality content is what you are looking for, Google search is still the best way.

This just made me think that the point behind the invention of AI was to save people time by avoiding tedious tasks like data scraping. People (who are, or will, depend on a 9-5 as their main source of income) are scared that they might lose their jobs to AI. But who knew that it can affect technologies like Google as well...

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Skroob]

It's a great tool. I still prefer finding authoritative sites and reading their content though. I especially miss an actual opinion and people pinning their flag to a mast.

Something I found playing around with it last night. I used it to find a reference to a biblical story I vaguely remembered, and it found it easily. Then I got into a "discussion" (as much as you can use that term, I guess) with it about the meaning of the story. It had an interpretation, which didn't make any sense based on the narrative of the story, and I spent some time trying to get it to understand where it was going wrong. It would acknowledge that I was correct, and then repeat the same interpretation anyway.

My reaction was that this thing is great for finding information, but not for interpretations of that information. I mean, it's the Bible. A book famous for every chapter and word having multiple, conflicting interpretations, and it gives me one and sticks to it.

However, now that I read your post... I've absolutely had "discussions" with actual humans that have gone that exact same way. Maybe this thing is better than I thought.

 

[Xeon]

ChatGPT is just a taste of a "monster" GPT-4 says Gary Marcus

With ChatGPT, OpenAI is currently testing a dialog-based general-purpose language model. According to cognitive scientist Gary Marcus, ChatGPT is just a foretaste of GPT-4.

the-decoder.com

Good. Show us what this monster will be like!

 

[Arrived2015]

How about this one:

https://twitter.com/i/web/status/1598513757805858820

View: https://twitter.com/tqbf/status/1598513757805858820

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Arrived2015]

Amazed at the speed and creativity of ChatGPT!

I used it today and this is the result:

Me: Story of a Christmas card that comes to life for two days

ChatGPT: Here is a story about a Christmas card that comes to life:

It was Christmas Eve, and Emily and her family were sitting around the tree, enjoying each other's company and exchanging gifts. As Emily was opening her last present, she noticed something strange about the Christmas card that her aunt had given her. It seemed to be moving on its own!

At first, Emily thought she was just seeing things, but as she looked closer, she realized that the characters on the front of the card were actually moving. She couldn't believe it! She showed the card to her family, and they were just as amazed as she was.

As they watched, the characters on the card came to life and stepped out of the paper. It was like something out of a fairy tale! The characters were all dressed in festive Christmas outfits and introduced themselves as Mr. and Mrs. Claus, Rudolph, and the elves.

Emily and her family were in awe as they talked with the characters and learned about their lives in the North Pole. They laughed and sang Christmas carols together, and the time flew by.

Before they knew it, the clock struck midnight, and the characters on the card began to fade back into the paper. Emily's aunt told her that the card had been enchanted by a special Christmas magic, and that it would only come to life for two days each year – Christmas Eve and Christmas Day.

As the characters disappeared, Emily and her family were filled with a warm and happy feeling, knowing that they had just experienced something truly magical.

They couldn't wait to see what the card would bring to life the next year.

I can see how I can use this tool as a starting point for writing fictional stories. All it needs, is expanding the paragraphs, add some images then package it up for selling.

Can see it becoming very, very addictive in 2023 (and beyond) especially with each updated version.

 

[GPM]

Is anyone else having issues getting ChatGPT to work? It has been 2 days now that I cant get a reply. I just keep getting errors: "An error occurred. If this issue persists please contact us through our help center at help.openai.com."

 

[Arithen]

Is anyone else having issues getting ChatGPT to work? It has been 2 days now that I cant get a reply. I just keep getting errors: "An error occurred. If this issue persists please contact us through our help center at help.openai.com."

Using it on the phone has more errors than on a pc (I’ve found). I’ve also found that refreshing the page every time there’s an error fixes it until I visit another tab or keep the window open for too long or ask it too many questions in a short period of time.

Having a multi-day span where you’re getting nothing but errors sounds like a bug you’d need to report. Try a different browser/device?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GPM]

Using it on the phone has more errors than on a pc (I’ve found). I’ve also found that refreshing the page every time there’s an error fixes it until I visit another tab or keep the window open for too long or ask it too many questions in a short period of time.

Having a multi-day span where you’re getting nothing but errors sounds like a bug you’d need to report. Try a different browser/device?

The error stopped on my PC, but everytime I try to use it on my laptop I get that error. I can't seem to get it to go away on the Laptop. Not sure why, I have not been able to figure it out. Resetting the computer, private browsers, turning on or off the VPN, nothing changes it.

I saw some posts that show how ChatGPT is going woke. I have not tried it myself, but I guess it can't even bring you some information on humans increasing their quality of life through burning fossil fuels or some such. Same with getting it to send you messages in the style of certain people. Those topics go against the narrative so they may be slowly being removed

 

[Andy Bell]

Update: Been running chatgpt like crazy recently to add some charts to articles, and I have to say, this thin gives some of the most inaccurate worst information I have ever seen, and it's insanely confident, which makes it hard to discern. It speaks like a confident 18-year-old trying to bullshit you on an exam. In the electrician niche, almost 90% of the information is wrong and dangerous, giving me info for how many devices I can put on a circuit that could burn your house down.

It's an amazing proof of concept, but I feel it could be more dangerous than good at this point with all the false facts it gives and how confident it makes it sound. We were about to get an internet flood of bs, and wrong information, the likes of which we had never seen before.

 

[BizyDad]

So I got writers block.

I asked it to help me write a eulogy for my best friend's funeral.

It gave me excellent tips. That was cool. It definitely saved me from having to do multiple Google searches. Within a minute, it got me thinking about 6 points I wanted to make in the eulogy. I was able to create an outline in my head.

Then I asked it to write it for me. Literally that's all I said.

It wrote 6 paragraphs, showed the right structure, included suggestions for light hearted vs serious content, and hit on 4 out of my 6 points.

All I have to do is add two paragraphs right in the middle and this thing is done.

But what's creepy is I thought that these four points were unique to my friend. I mean, I haven't had another friend like her. It's hard to believe that an AI would aggregate four different fairly unique positive attributes about a person. Is everyone's best friend a super intuitive altruistic person?

Its easier to believe that Elon Musk hired aliens to implant a nuerolink when I wasn't looking, because this thing was just too good.

It's crazy to me that I literally asked it two things, all I gave it was two simplistic questions, and it spit out a better speech than I had been able to achieve in the last 30 minutes.

I saw some posts that show how ChatGPT is going woke. I have not tried it myself, but I guess it can't even bring you some information on humans increasing their quality of life through burning fossil fuels or some such. Same with getting it to send you messages in the style of certain people. Those topics go against the narrative so they may be slowly being removed

Yeah, the co founder apparently runs an solar company and electric car company. I think he's super woke.

Update: Been running chatgpt like crazy recently to add some charts to articles, and I have to say, this thin gives some of the most inaccurate worst information I have ever seen, and it's insanely confident, which makes it hard to discern. It speaks like a confident 18-year-old trying to bullshit you on an exam. In the electrician niche, almost 90% of the information is wrong and dangerous, giving me info for how many devices I can put on a circuit that could burn your house down.

It's an amazing proof of concept, but I feel it could be more dangerous than good at this point with all the false facts it gives and how confident it makes it sound. We were about to get an internet flood of bs, and wrong information, the likes of which we had never seen before.

What troubles me is the thought that enough people/bots can post this false info, thus training future generations of AI on false info.

When someone first said we live in a post truth world, I thought it was hyperbole. But we get closer and closer...

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Kak]

Is anyone else having issues getting ChatGPT to work? It has been 2 days now that I cant get a reply. I just keep getting errors: "An error occurred. If this issue persists please contact us through our help center at help.openai.com."

Just refresh the page.

 

[TheCj]

Update: Been running chatgpt like crazy recently to add some charts to articles, and I have to say, this thin gives some of the most inaccurate worst information I have ever seen, and it's insanely confident, which makes it hard to discern. It speaks like a confident 18-year-old trying to bullshit you on an exam. In the electrician niche, almost 90% of the information is wrong and dangerous, giving me info for how many devices I can put on a circuit that could burn your house down.

It's an amazing proof of concept, but I feel it could be more dangerous than good at this point with all the false facts it gives and how confident it makes it sound. We were about to get an internet flood of bs, and wrong information, the likes of which we had never seen before.

The confident 18yr old analogy is spot on!

Playing with chatgpt I came to similar conclusion.

The line that comes to mind is, getting data is great, getting your own data is the best. With this chatgpt, the data being fed into it and the biases/agenda is what is concerning.

Just like youtube removing dislikes so people can't easily tell, if someone is showing something dangerous, wrong, unsafe, biased. So to is it hard for someone who doesn't know any better to see the biases, and know the agenda behind answers.

It feels very much like, the truman show. Your free! just don't go walking into the room wall/boundary that can't get out of. But turn around and look here it's amazing and your free!

 

[Andreas Thiel]

Yeah, the co founder apparently runs an solar company and electric car company. I think he's super woke.

Elon seems to cut all ties with OpenAI. He stated that when he cofounded it it was both Open Source and non profit, both are no longer true.

Twitter used to be data source for OpenAI, Elon stopped that. Don't know much about Brockman and Altman, but all AI systems seem to be extremely woke.

Has anybody played with MidJourney? Every other prompt seems to lead to a "do you want to get banned?" reply for me. It is a serious issue.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[GPM]

Elon seems to cut all ties with OpenAI. He stated that when he cofounded it it was both Open Source and non profit, both are no longer true.

Twitter used to be data source for OpenAI, Elon stopped that. Don't know much about Brockman and Altman, but all AI systems seem to be extremely woke.

Has anybody played with MidJourney? Every other prompt seems to lead to a "do you want to get banned?" reply for me. It is a serious issue.

I was playing with midjourney for a while, but ended up getting frustrated. Anything had may have text in it, or human hands gets absolutely butchered. Faces on books and stuff are ruined and distorted as well.

For fantasy images or landscapes and stuff it is absolutely insane though. Works building and landscaping and drawing fantasy creatures and such it is just crazy good.

I have also seen it build some pretty cool simple logos as long as you keep text out of it, or keep it to 1 or 2 letters.

And @Kak thanks for suggesting I refresh the page, I never would have thought of that!! Haha /s

 

[Andy Bell]

The huge risk too is a massive amount of writers are now turning to it to autogenerated text which means when it scrapes the web again to learn it's going to be scraping its own chatgpt results essentially training itself on its own wrong information making future iterations a complete mess of innacuracy.

 

[Kevin88660]

R

The error stopped on my PC, but everytime I try to use it on my laptop I get that error. I can't seem to get it to go away on the Laptop. Not sure why, I have not been able to figure it out. Resetting the computer, private browsers, turning on or off the VPN, nothing changes it.

I saw some posts that show how ChatGPT is going woke. I have not tried it myself, but I guess it can't even bring you some information on humans increasing their quality of life through burning fossil fuels or some such. Same with getting it to send you messages in the style of certain people. Those topics go against the narrative so they may be slowly being removed

Refresh or change broswer

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Ing]

now that is something I don t understand. When I asked about dayly information, that was answered :

I'm sorry, but my knowledge is limited to what was known up until my knowledge cutoff in 2021. I do not have access to any information beyond that point in time. Is there anything else I can help you with using the information that I have?

 

[CoolBeans]

Found an interesting use case for ChatGPTtoday: translating old French, Spanish and Italian books that were not translated in English.

I wonder if anyone will find any value in it though.

 

[Andreas Thiel]

I was playing with midjourney for a while, but ended up getting frustrated. Anything had may have text in it, or human hands gets absolutely butchered. Faces on books and stuff are ruined and distorted as well.

For fantasy images or landscapes and stuff it is absolutely insane though. Works building and landscaping and drawing fantasy creatures and such it is just crazy good.

I have also seen it build some pretty cool simple logos as long as you keep text out of it, or keep it to 1 or 2 letters.

And @Kak thanks for suggesting I refresh the page, I never would have thought of that!! Haha /s

Yeah it is interesting what Midjourney is bad at compared to what it absolutely nails. Something worth trying is adding a link to a photo in the prompt. My new avatar is something that Midjourney created based on a photo of mine. I added "undercut" as a keyword and it managed to keep a lot of the likeness intact. Ear in the background is one of those disgusting accidents, though ^^
As long as it can stick to the original photo it does okay. You can't tell it to add people or other crazy things like that. You can add a rough collage in Photoshop or Gimp and see if Midjourney can fix it. Got a few nice image out of that approach, especially when you add a painter name to get the desired style.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[WillHurtDontCare]

@MJ DeMarco can we include AI text prompt tools that aren't chatgpt tools in this thread? I've had fun with Midjourney

you make these in a discord server and the free plan images are public - so you can see what other people come up with

also, I'm starting to believe that this AI wasn't built on software & math but instead was built using whatever sorcery CERN is using to try to open a portal to hell

horror comic style mcdonalds employees running from a monster ronald mcdonald

a crusader holding a bible and an assault rifle

 

[MJ DeMarco]

@MJ DeMarco can we include AI text prompt tools that aren't chatgpt tools in this thread? I've had fun with Midjourney

you make these in a discord server and the free plan images are public - so you can see what other people come up with

also, I'm starting to believe that this AI wasn't built on software & math but instead was built using whatever sorcery CERN is using to try to open a portal to hell

horror comic style mcdonalds employees running from a monster ronald mcdonald

a crusader holding a bible and an assault rifle

I'd say use this thread.

IDEA - ChatGPT / AI / MidJourney Creation Repository: Share Your Useful Creations

I am aware there are two or three ChatGPT-related threads on the forum. However, I wanted to make this one separate, to highlight and share random tools and things we create using the AI. Especially for non-coders using it to create neat little things for business, productivity and so on. While...

www.thefastlaneforum.com