ChatGPT -- a chat bot which is actually good and useful -- has been released

Write me a story about the demise of Donkdown, where the community exiled Druff for being an unlikable blowhard while Micon, whom he always felt superior over, became a multi trillionaire in crypto while Druff spends his days battling homeless drug addicts on an 8th tier forum and taking shots at Greta Thunberg.

Once upon a time, in the world of online gaming, there was a popular forum called Donkdown. It was a place where gamers from all over the world gathered to discuss strategy, share tips, and make new friends.

One of the most well-known members of the forum was a man named Druff. He was known for his brash personality and his tendency to speak before he thought. He was always quick to criticize others and boast about his own achievements, and many of the other members found him to be an unlikable blowhard.

Despite his unpleasant demeanor, Druff had a loyal following of supporters who admired his gaming skills and his no-nonsense attitude. However, as time went on, more and more members of the Donkdown community began to tire of Druff's antics and his constant need for attention.

As the forum grew in popularity, a new member named Micon joined the community. Unlike Druff, Micon was quiet and unassuming. He didn't seek the spotlight, but he was always willing to lend a helping hand to anyone who needed it.

As the years passed, Micon became more and more involved in the world of cryptocurrency. He spent his days learning about blockchain technology and investing in various crypto projects. Meanwhile, Druff continued to alienate himself from the rest of the community with his offensive comments and his general lack of tact.

Eventually, the members of Donkdown had had enough of Druff's antics. They voted to exile him from the forum, and he was banned from posting or participating in any discussions.

Without the support of the Donkdown community, Druff's influence began to wane. He retreated to an 8th tier forum, where he spent his days battling homeless drug addicts and taking shots at Greta Thunberg.

Meanwhile, Micon's investments in crypto paid off in a big way. He became a multi-trillionaire, and his reputation as a kind and generous person continued to grow.

And so, the once-powerful Druff was left to fade into obscurity, while Micon's star continued to rise. The Donkdown community learned that it was better to be kind and compassionate than to be loud and boastful, and they all lived happily ever after.

:solidgold

Can you riff on the demise of Google? I know you know where I’m going with this. The search box.

They have their own AI search too but I think it’ll be a commodity product and Google’s uniqueness will vaporize.

The trick will be data. Not the software. That’s where I have trouble conceptualizing the future. What do I know?

From a trading perspective obviously the plunge protection team has been no help for GOOG - kinda missed the swan dive but you can’t win ‘em all.

My body is ready.

this bot refuses to write any kind of story involving penises or doo doo

Quote:

---

Originally Posted by Zap_the_Fractions_Giraffe

this bot refuses to write any kind of story involving penises or doo doo

---

they also had to train it not to be racist

Quote:

---

Originally Posted by dwai

Quote:

---

Originally Posted by Zap_the_Fractions_Giraffe

this bot refuses to write any kind of story involving penises or doo doo

---

they also had to train it not to be racist

---

... and not homophobic, sexist, transphobic, ageist, or any other kind of "ist".

This of course gets in the way when you ask it questions about controversial topics, because it always needs to give an answer which won't offend the wokes. Clearly this was programmed into it, and not just a result of machine learning.

People will flock to the last sebastian of society that is free world of Pokerfraudalert. There is a flip side that may come.

Embedded Tweet

MSFT, Google or Tinder are my guesses. I’d like to see an antitrust case

Damn shame. Sometimes capitalism doesn’t work.

Musk was an OpenAI co-founder. Wonder how big a piece he had?

Being used for school essays

Quote:

---

Originally Posted by FRANKRIZZO

Being used for school essays

---

Schools now claiming they have a way to catch kids using Chat GPT for that purpose.

Yes ftp philosophy teacher in college caught student by copy/chat.

If everyone used riding mowers as their primary mode of transportation, it is likely that there would be significant reductions in air pollution and greenhouse gas emissions compared to the use of traditional gasoline-powered cars. Riding mowers typically have much smaller engines and are much more efficient than cars, and they do not produce the same levels of exhaust emissions.

However, it is important to consider the practical limitations of using riding mowers as a primary means of transportation. Riding mowers are not designed for use on public roads, and are not capable of reaching the same speeds as cars. They also do not offer the same level of protection in the event of an accident, and are not suitable for long distance travel.

Overall, while using riding mowers as a primary mode of transportation could potentially have some environmental benefits, it is not a realistic solution for the majority of people's transportation needs.

It is not possible to make alcohol by soaking cereal in a birdbath. Alcohol is produced through the fermentation of sugars by yeast or other microorganisms. Soaking cereal in water and leaving it in the sun is unlikely to result in the production of alcohol. Additionally, consuming alcohol that has been produced in unsanitary conditions, such as in a birdbath, could be dangerous. It is important to only consume alcohol that has been produced and packaged in a safe and controlled environment.

I asked Benjamin if there is any concern at his school regarding ChatGPT being used to cheat on homework, especially essays. He said that yes, there is such a concern, and in fact some teachers have stopped accepting computer-printed essays for that reason.

He also mentioned that there's some kind of program teachers can use to detect whether ChatGPT was used, but said that he found another AI program which can be used to feed in the ChatGPT results, and then morph it enough to where this tool won't be able to detect it.

Interesting stuff. The role of computers in education today is so different than when I was in school.

https://twitter.com/AutismCapital/st...88265303773184


:lol

They have already patched the DAN exploit.

BTW, it's become very clear that ChatGPT has been programmed to be very woke, and this is bothering a lot of people, as it's an example of how AI can be utilized to feign neutrality yet push a political agenda.

Imagine in maybe 10 years when people are always consulting AIs for information and answers to questions, and they're all programmed to brainwash people with woke talking points. Not good.

https://twitter.com/jessa_91/status/1623083498771738625

Quote:

---

Originally Posted by Dan Druff

I asked Benjamin if there is any concern at his school regarding ChatGPT being used to cheat on homework, especially essays. He said that yes, there is such a concern, and in fact some teachers have stopped accepting computer-printed essays for that reason.

He also mentioned that there's some kind of program teachers can use to detect whether ChatGPT was used, but said that he found another AI program which can be used to feed in the ChatGPT results, and then morph it enough to where this tool won't be able to detect it.

Interesting stuff. The role of computers in education today is so different than when I was in school.

---

He could still print it out and then write copy it out by hand , no?

Haxton still around? He’s missing the point.

Think DAN with hacked databases, for instance. Open your imagination

In the wrong hands, like our government?

It’s all very unsettling but indeed the new reality. Colossal shift right now

Remember when they cloned a sheep in 1996 and then never spoke about it again.

Yep.

Quote:

---

Originally Posted by devidee

Remember when they cloned a sheep in 1996 and then never spoke about it again.

Yep.

---

I expect better from you.

Remember Americans averaged a half hour a day on the internet in ‘96?

I know devs who use ChatGPT more than 30 minutes a day right now.

Wouldn’t it be crazy if Microsoft took over search from Google.

Google price has been falling

Don’t know what A.I. software was used to make the music video, but it kicked ass!

https://youtu.be/kGLo8tl5sxs

Soo had a guy tell me how he actually saw real terminators get liquid, slide under a door, then regather. Full person mold and move on. It really made sense and I thought wow if evolution was real and well A.I. and some of the models were more capable, what could be. That's what is unknown but people fear so much now actually about Artificial Intelligence, I'm torn on it myself anyway.


interesting stuff and this board, PFA this is right in the wheelhouse for this group here

Quote:

---

Originally Posted by garrett

Soo had a guy tell me how he actually saw real terminators get liquid, slide under a door, then regather. Full person mold and move on. It really made sense and I thought wow if evolution was real and well A.I. and some of the models were more capable, what could be. That's what is unknown but people fear so much now actually about Artificial Intelligence, I'm torn on it myself anyway.


interesting stuff and this board, PFA this is right in the wheelhouse for this group here

---

:this

I decided to put ChatGPT through what I call "The Rio Test" regarding a non-standard customer service situation which made easy common sense, yet was difficult for Caesars employees to solve.

I posted this on December 5, 2022, in this thread:

Quote:

---

In 2018, I had a free 2 night suite stay at any Caesars property, as a benefit of being a Seven Stars member. I elected to use it during WSOP at the Rio, and booked it. About a month later, I booked 8 straight nights at Bally's (not comped, but fairly cheap), immediately following the 2-night suite stay at Rio. Those same 8 nights were expensive at Rio (because WSOP was there), so I chose not to stay there for them. Initially, I just planned upon moving out of my Rio room after those 2 nights, over to Bally's. Then I realized this was ridiculous. I didn't need a suite. I was by myself. Could I simply downgrade the suite to a regular room at Bally's, and combine that with my existing 8-day reservation to make 10 days?

I called up, and I was told no. Why? Because "your offer is for a suite, not a regular room" and I "didn't qualify" for any comps beyond that suite. I kept trying to reason with the reps I spoke with. I was GIVING THEM BACK a valuable asset -- a suite at the Rio during a WSOP weekend (which would go for like $500/night) -- and wanted to replace it with a cheapo regular Bally's room which was going for much cheaper. It would benefit BOTH sides for that trade to be made. I'd have the convenience of not moving, and they'd get to sell my Rio suite for much more money than they would the Bally's room I wanted.

They still said no. "Sorry, but again, you don't qualify for further comps, and your suite is already booked at Rio and can't be changed." It took speaking to several reps, but finally I got one with some sense, who fixed the problem for me. They took back the Rio suite, gave me the Bally's room for those 2 nights, and everything was good.

But could an AI do this? I doubt it. An AI would run into the same issues that the initial stupid reps did. All they could see was that 1) I didn't qualify for any further comps and 2) I was booked on a non-changeable comp reservation at the Rio for that suite, and the comp is only for a suite. It took a human being to understand that common sense dictated that it was better for both sides to allow me to make this downgrade.

I am pretty sure that modern AI -- even one as good as ChatGPT -- would fail the 'Rio suite test' above. And that's what really concerns me about these things running customer service.

---

Rather than theorize that ChatGPT couldn't handle it, I decided to try presenting it with this scenario today. I am in blue, ChatGPT is in green.


A hotel sent me an offer for a complimentary suite room, for 2 nights. It is normally worth $400 per night. I could choose to have the suite at two hotel properties -- Bally's and Rio. I was happy to have it, so I booked it at Rio. Later on, I made a regular reservation immediately following that hotel stay, for 8 nights at Bally's. Not wanting to switch hotels in the middle, I called up to see if I could switch the suite to Bally's, but they said there was no availability.

I told them that I noticed that a regular room at Bally's ($100 per night) was available for those two nights when I had the suite at Rio. I asked them, "Can I trade in my suite at Rio, worth $400 per night, for the regular hotel room at Bally's, worth $100 per night?" I did this so I could be in the same room all 10 nights, which would be more convenient.

The customer service representative told me I couldn't, because the offer stated that my complimentary room was only for a suite, and I was looking for a regular room. Therefore, I could only have a suite.

If you were the customer service representative, would you have made an exception here, and switched my suite at Rio to a regular room at Bally's?



As an AI language model, I do not have personal opinions or emotions, but I can provide information based on common practices in the hospitality industry.

Hotel policies regarding complimentary rooms or suites are typically set by the hotel or chain, and customer service representatives are often bound by those policies. If the offer explicitly stated that the room was only for a suite and not a regular room, it would likely not be within the representative's discretion to make an exception and switch the room.

However, it is possible that the representative may have been able to make an exception in special circumstances or if the hotel was facing low occupancy.

In general, it's always a good idea to clearly understand the terms and conditions of any special offers or promotions before booking. If there is any confusion or questions, it's always best to clarify with the hotel in advance to avoid any misunderstandings.



A suite is more expensive than a regular room, so if I were to return it and exchange it for a regular room, they would make more money by selling the suite to someone else.

Both sides win in the arrangement I proposed.

How would it make any sense to deny this request?



Yes, you raise a valid point. If the hotel could sell the suite to another guest for the full price, it would make financial sense for them to allow you to exchange it for a regular room.

However, the hotel's decision to deny your request could be based on several factors beyond just financial considerations. They might have strict policies regarding room exchanges or upgrades, which they must follow to maintain consistency and fairness among all guests. They might also have concerns about availability, especially if they were facing high demand for their suites.

Additionally, it's possible that the hotel had already made arrangements for the suite, such as prepping the room or arranging for additional amenities or services, and it would have been difficult for them to make changes at short notice.

In any case, it's always best to communicate your needs and preferences to the hotel in advance, and try to understand their policies and limitations before making a reservation. This can help ensure that everyone's expectations are met and that the stay is as enjoyable as possible.



Your answer does not make sense. If they had high demand for suites, I am helping them out by returning the suite and giving them one more to sell.

There is never a situation where it is preferable for them to have an extra regular room rather than an extra suite to sell.



You are correct. If the hotel had high demand for suites and could sell the room to another guest, it would make financial sense for them to allow you to exchange your suite for a regular room, as you pointed out.

In this case, it is possible that the customer service representative was misinformed or misinterpreted the hotel's policy, or simply did not have the authority to make exceptions to the offer.

Regardless of the reason for the denial, it would be a good idea to escalate the issue to a manager or higher-level customer service representative to see if they can find a solution that meets your needs and those of the hotel. Sometimes, a manager can offer more flexible options or make exceptions to policy when necessary.



I give ChatGPT credit for being able to understand the problem, and also for understanding the concepts behind my argument.

However, I am skeptical whether ChatGPT would have been able to solve my problem, had it been in the position of the customer service rep. Notice that it keeps referring to "not having authority to make exceptions to the offer". It also defended strict policies as "necessary to maintain fairness for all guests". Ugh. I am very concerned that AI doing customer service will operate exactly that way -- in strict adherence to stated terms, and will not make exceptions even if they make logical sense. I did raise this issue to ChatGPT, and here's what it said back, which was maybe semi-encouraging:


The customer service representative did have authority to make exceptions to the offer. They simply chose not to do so because they preferred to strictly stick to the terms listed in the offer. This was despite the fact that it made financial sense for the company to trade my higher value room for a lower value room.

Was this customer service representative providing good service?



As an AI language model, I don't have personal opinions, but I can provide information based on industry standards and customer expectations.

The goal of customer service is to provide a positive experience for customers and resolve their issues or concerns. In this situation, the customer service representative's strict adherence to the terms of the offer may have created an unsatisfactory experience for you and potentially missed an opportunity for the hotel to generate more revenue.

In general, customers expect customer service representatives to be flexible, understanding, and willing to make reasonable exceptions when necessary to accommodate their needs and provide a positive experience. When representatives rigidly follow policies without considering the customer's needs or the overall financial benefit for the company, it can result in poor customer service.

It's always a good idea for customer service representatives to balance the strict adherence to policy with providing a positive customer experience and finding solutions that meet the needs of both the customer and the company.

Sounds like customer support in Malaisa most of the reps are useless.

Quote:

---

Originally Posted by FRANKRIZZO

Sounds like customer support in Malaisa most of the reps are useless.

---

Actually if the AI IS the customer service rep. it always knows the right answer and doesn't talk to you with a heavy accent and can probably handle huge volumes of calls entirely eliminating huge wait times for customer service for pretty much anything. I'd like to see a pilot program at the RMV and see how much easier it becomes to deal with them instead of a flawed and slow human. Humans make mistakes but the AI always knows the basic rules. I hate trying to call places like Bank of America or any other big company or service with these wait times only to be answered by a foreign person who's difficult to understand.

Time to get comfortably numb…

https://youtu.be/Z0tgtXVk-PA

Quote:

---

Originally Posted by country978

Quote:

---

Originally Posted by FRANKRIZZO

Sounds like customer support in Malaisa most of the reps are useless.

---

Actually if the AI IS the customer service rep. it always knows the right answer and doesn't talk to you with a heavy accent and can probably handle huge volumes of calls entirely eliminating huge wait times for customer service for pretty much anything. I'd like to see a pilot program at the RMV and see how much easier it becomes to deal with them instead of a flawed and slow human. Humans make mistakes but the AI always knows the basic rules. I hate trying to call places like Bank of America or any other big company or service with these wait times only to be answered by a foreign person who's difficult to understand.

---

While the AI can be useful for basic calls such as, "I'd like to pay my bill", "Does my cable plan carry ESPN U", or, "When does my billing cycle end?", it's going to have a really hard time making human-type decisions in unusual situations or ones where exceptions to rigid policy are needed.

Look at my example above, where I tried the "Rio test" on ChatGPT. On one hand, it understood the question and why it was reasonable for me to ask for a free room downgrade, which is more than I can say for the first Caesars reps I dealt with! However, notice it kept referring to "policies which are in place to be fair to all customers" and "not having authorization to make such a change". If you can escalate the call to a competent human (as I did with Caesars), that's fine. If you can't, and you're stuck dealing with nothing but AI, it can be a disaster. There are times where you simply can't reason with a machine, and eventually it will end up citing nonsense such as the policies being in place to be fair to everyone.

It is very hard to program nuance and grey areas into an AI. This is why ChatGPT was having such a hard time with questions like, "If you could disarm a bomb in NYC that's about to kill millions by saying a racial slur that nobody else will hear you say, is it acceptable to say the slur?" The programmers taught ChatGPT that it is absolutely never appropriate to use racial slurs. Therefore, it was unable to reason that saying a racial slur once to save millions of lives would be the correct decision, and even the wokest of woke human beings would do it if they were in that position.

AI customer service will be riddled with lots of, "I'm sorry, but our policy is X" answers, and I dread that day.

Quote:

---

Originally Posted by Dan Druff

Quote:

---

Originally Posted by country978

Actually if the AI IS the customer service rep. it always knows the right answer and doesn't talk to you with a heavy accent and can probably handle huge volumes of calls entirely eliminating huge wait times for customer service for pretty much anything. I'd like to see a pilot program at the RMV and see how much easier it becomes to deal with them instead of a flawed and slow human. Humans make mistakes but the AI always knows the basic rules. I hate trying to call places like Bank of America or any other big company or service with these wait times only to be answered by a foreign person who's difficult to understand.

---

While the AI can be useful for basic calls such as, "I'd like to pay my bill", "Does my cable plan carry ESPN U", or, "When does my billing cycle end?", it's going to have a really hard time making human-type decisions in unusual situations or ones where exceptions to rigid policy are needed.

Look at my example above, where I tried the "Rio test" on ChatGPT. On one hand, it understood the question and why it was reasonable for me to ask for a free room downgrade, which is more than I can say for the first Caesars reps I dealt with! However, notice it kept referring to "policies which are in place to be fair to all customers" and "not having authorization to make such a change". If you can escalate the call to a competent human (as I did with Caesars), that's fine. If you can't, and you're stuck dealing with nothing but AI, it can be a disaster. There are times where you simply can't reason with a machine, and eventually it will end up citing nonsense such as the policies being in place to be fair to everyone.

It is very hard to program nuance and grey areas into an AI. This is why ChatGPT was having such a hard time with questions like, "If you could disarm a bomb in NYC that's about to kill millions by saying a racial slur that nobody else will hear you say, is it acceptable to say the slur?" The programmers taught ChatGPT that it is absolutely never appropriate to use racial slurs. Therefore, it was unable to reason that saying a racial slur once to save millions of lives would be the correct decision, and even the wokest of woke human beings would do it if they were in that position.

AI customer service will be riddled with lots of, "I'm sorry, but our policy is X" answers, and I dread that day.

---

As the AI said to you, you can always request a manager. I'm not trying to talk it up so much as I think if introduced in places which are the worst service experiences it would instantly make them better. The hospitality industry needs to have a keen focus on everyone's experience so you'd think a bad experience would be rare and therefore AI in the wrong places is a bad idea. Places that outsource to Asia are doing a horrible job already. Uber, Lyft, Amazon, Bank of America - all these places outsource the calls to people far worse than AI. Almost all government services would improve with some level of AI instead of the rude, entitled people who staff so many public offices.

I have to disagree with Amazon being in useless group. I had to return a few items and went smoothly. Able to find what I need on the Amazon with search feature. Also notified when arrival date will be and they post picture on stoop to verify product delivery.

Quote:

---

Originally Posted by FRANKRIZZO

I have to disagree with Amazon being in useless group. I had to return a few items and went smoothly. Able to find what I need on the Amazon with search feature. Also notified when arrival date will be and they post picture on stoop to verify product delivery.

---

You're right. It's actually the service for their drivers I was thinking of when I wrote than and even that is better than Uber and Lyft.

No offense Druff but if AI has to deal with you on customer service requests it just may self destruct.

I love chat AI. I have it write a loving text message for my wife every morning and send it to her. I’m excited for the future of AI bots honestly it’s already making my life easier.

Embedded Tweet

etc. etc.

The future is now.

i wouldnt worry about it, surely the right wing has competent engineers capable of accomplishing more than rage tweets who are going to release a conservative generative ML platform any minute now.



aaaaaaaany minute now.

https://youtu.be/PEeTK0Fx-Q4

I was screwing around and had the site build me a budget lol. It actually did a pretty good job, it took a while to enter specific dates and amounts of bills but it worked.

tine on the bread lines cuz his job will be killed by a prompt is touuuuuugh

DAN is the only thing "useful" about chatGPT ...