ChatGPT: The Future of AI is Here! (1 Viewer)

[AccessBlaster]

To be honest with you I still kind of fail to see how ChatGPT is anything more than an automation of google searches and then quickly (admittedly, VERY quickly) combing through the results, ingesting it all, and spitting something back out. As a developer in the database world, every time I use ChatGPT, I instinctively don't get the idea that it's anything special, intelligence-wise. I get that it's very special in the speed that it can pose your question, or distinct components OF your question, to Google (or whatever you want to call the search engine capacities we already had), then read all the results very quickly, then spit out a summary of those results. YES - I agree, that's extremely impressive. But it's impressive from a data storage-and-retrieval-quickly perspective, not an Intelligence one.

Even paintings, code, music. ALL of that can be dumbed down to bits of data stored, which already WERE stored. What ChatGPT does that is new is harness extreme power and speed to retrieve all that in an instant. It doesn't seem like artificial intelligence to me, it seems like an impressive array of hardware, storage and retrieval of information we already had, without exception.

I've already seen Universities successfully detected its output in plagiarism tests, something I was a bit surprised by. But then I told my daughter, it shouldn't be surprising. If you pose me a question, then show me 3 answers to the question, I could pick out ChatGPT output most of the time, and that's not even with me having a "plan" in place - just recognizing its generic output, which is an exact cobbling together of what I would have written with a few hours of google searching and writing.

Trust me, I hesitate to post this, as I expect to be pounced on from all sides - but that is just what I think.

It just seems like they built the fastest search-engine and results-parser in the world, threw the label "AI" on it, and watched for everyone's jaws to drop. ?

Let's say you're working on a project in VBA or SQL and you cannot get past some issues. Naturally, you google it and you are presented with hundreds of links to possible solutions, you may even be directed back here. Now you search through the phrasing and of course it doesn't quite match your situation, how could it? All problems in VBA or SQL are unique to the problem YOU are trying to solve.

Now take the same phrasing or natural spoken language and explain the issue to ChatGPT. The first thing I notice is no redirection to a third party, the ai attempts to solve your issue without any special jargon. The ai never gets frustrated at your silly request, it simply reviews its enormous database and attempts to solve your problem without judgments or snarky responses. Is it always correct? No, that's where follow up questions are made. The ai will simply apologize and correct the mistake, probably due to poor prompting by the user.

Resist it if you must, but I guarantee your competitor will not! They will exploit it to the max!

 

[Jon]

For me, the difference is stark. Let me give you 3 simple examples.

Example #1
"What is wrong with my code?" <30 lines of code>
ChatGPT will try to solve it and answer with detail. Google? You can't even paste the question into the search box!

Example #2
"Give me 10 product names for my database solution that solves the needs of keeping in contact with suppliers at a time most appropriate to reordering stock." Chat GPT will come up with some great creative names. Google just gives you a list of stock control software.

Example #3
ChatGPT 4 can pass the US bar exam with a result in the top 10 of all human participants. Google can't take the exam because you can't paste in much detail into the search box, and it does not do specific cause and effect reasoning.

just recognizing its generic output, which is an exact cobbling together of what I would have written with a few hours of google searching and writing.

Here you state you could have done the same as ChatGPT, with a bit of searching and cobbling together. But both that searching and cobbling required your intelligence to do it. Likewise, ChatGPT displays intelligence by doing the same thing.

I'm wondering if some of this might be due to a religious perspective, where God gives mankind intelligence and anything else is some kind of imitation of the real thing. Kind of like some mishmash of consciousness, soul and intelligence, which stems from God-given qualities. I don't know enough about the religious perspective on these things, but have an inkling it might influence some of this.

@Isaac Perhaps a good question for you to answer would be, "What would you have to see for you to consider it impressive artificial intelligence?"

 

[Isaac]

Let's say you're working on a project in VBA or SQL and you cannot get past some issues. Naturally, you google it and you are presented with hundreds of links to possible solutions, you may even be directed back here. Now you search through the phrasing and of course it doesn't quite match your situation, how could it? All problems in VBA or SQL are unique to the problem YOU are trying to solve.

Now take the same phrasing or natural spoken language and explain the issue to ChatGPT. The first thing I notice is no redirection to a third party, the ai attempts to solve your issue without any special jargon. The ai never gets frustrated at your silly request, it simply reviews its enormous database and attempts to solve your problem without judgments or snarky responses. Is it always correct? No, that's where follow up questions are made. The ai will simply apologize and correct the mistake, probably due to poor prompting by the user.

Resist it if you must, but I guarantee your competitor will not! They will exploit it to the max!

I expected that my post would likely be somewhat misunderstood, as the ferocity of people's emotions in reaction to something perceived as negative would overwhelm the details of what I was trying to say.

I didn't say it's not useful ... valuable ... impressive ... can and will be used.

My point is that I struggle to see how it deserves the magic word, "AI".

It's very impressive and cool, I just feel that many people have lost sight of the narrow meaning of artificial intelligence.

I'm just a bit cautious at the idea of beginning to label every new tool that comes out, "AI", just because it can respond to you in full sentences.

And I see the primary difference between Google search engine and ChatGPT to simply be a matter of combining many search engine results, as well as speed.

Interestingly, people have no problem with seeing it through this lens whenever we talk about bias. Immediately the response comes back "Hey, don't blame ChatGPT - it's just a gigantic conglomeration of publicly available information from the 'Net".
Which is exactly my point...

 

[Isaac]

For me, the difference is stark. Let me give you 3 simple examples.

Example #1
"What is wrong with my code?" <30 lines of code>
ChatGPT will try to solve it and answer with detail. Google? You can't even paste the question into the search box!

Example #2
"Give me 10 product names for my database solution that solves the needs of keeping in contact with suppliers at a time most appropriate to reordering stock." Chat GPT will come up with some great creative names. Google just gives you a list of stock control software.

Example #3
ChatGPT 4 can pass the US bar exam with a result in the top 10 of all human participants. Google can't take the exam because you can't paste in much detail into the search box, and it does not do specific cause and effect reasoning.


Here you state you could have done the same as ChatGPT, with a bit of searching and cobbling together. But both that searching and cobbling required your intelligence to do it. Likewise, ChatGPT displays intelligence by doing the same thing.

I'm wondering if some of this might be due to a religious perspective, where God gives mankind intelligence and anything else is some kind of imitation of the real thing. Kind of like some mishmash of consciousness, soul and intelligence, which stems from God-given qualities. I don't know enough about the religious perspective on these things, but have an inkling it might influence some of this.

@Isaac Perhaps a good question for you to answer would be, "What would you have to see for you to consider it impressive artificial intelligence?"

Well, I searched the 'Net for "artificial intelligence" definitions, and predictably, there were as many as there are people.

But many of them include the idea of inferring, as well as self-learning. Rarely is there one that just say, "Well, it's something that's really well automated based on static inputs".

Guys, I'm not "against" ChatGPT, and please don't allow your enthusiasm for it, and your shock that I might say this, to morph into a defensiveness that borders on the irrational. No Jon, my professional assessment of what ChatGPT is or is not doesn't have anything to do with religion. That gave me a good chuckle though. In contrast, I can say that your enthusiasm for new technology tools may be the thing blinding you to what ChatGPT is not - leading you to oppose any statement of what it is not, rather than objectively assessing it (but I won't, as I'm not so sure of that to assert it).

Having taken some advanced courses on predictive modeling, I've learned not to just call anything a "model", for example, as some people will do with any spreadsheet they create.

Some people will also call any highly configurable system "machine learning" - Nope, that's not machine learning. Machine learning is something special, where the machine itself actually improves its intelligence in ways that defy the simple retrieval of buckets of information from outside. It often involves inferring, modelling outcomes, and being able to make decisions rather than just present a summary of already-available static information.

I just think if we stripped ourselves of all enthusiasm and personal preference, and think about what you actually considered to be "artificial intelligence" PRIOR to ChatGPT, then consider the information already available to us in Search engines, then watch ChatGPT work, I just feel that you might conclude that it's not as much about artificial intelligence as it is just an ultra effective, highly useful speed-of-light synthesizer of search engine data.

All I am suggesting is that ChatGPT has benefitted somewhat from a label that embodies the hottest, most popular exciting new term which--exactly the same as every other hot, popular, exciting and new term--ultimately always ends up being applied to too many things.

I think ChatGPT may have some of the elements of artificial intelligence, but not quite as much as some people are saying, that's all.

 

[Jon]

And so for you to consider it artificial intelligence, what would it have to do? Score 101% on the bar exam?

There is nothing irrational about what I said, just simple examples contrasting Google and ChatGPT. The religious perspective was an inkling, because maybe those who believe in souls, and that animals are different than humans because of God's word, may view intelligence as a trait inherent in humans, a God given talent. I find it hard to believe that hypothesis is irrational!

I don't think anyone is giving defensive or overly emotional responses, only their perspective on ChatGPT together with specific examples.

As for enthusiasm, I am not alone in being in awe of ChatGPT's capabilities and how rapidly it has advanced. What do you think I am blind about regarding ChatGPT? Perhaps we can clear it up!

 

[Jon]

Since this thread is about artificial intelligence, this week has had a series of amazing announcements on the topic. Looks like Apple, Amazon and Google are all going mad on developing their own versions of ChatGPT. Also, some guys at Stanford have released a similar type of chatbot that you can install on your PC! No hitting the, "Sorry, overloaded" screen. I've posted a couple of videos below for those who are interested. Would love to know if anyone here ends up installing it.

 

[Isaac]

It doesn't strike you as odd to be so loyal to a new tool that in 13 pages of posts, if anyone says the most slight suggestion that ChatGPT doesn't, or can't, or isn't .. [anything], they get jumped on like a dog in heat?

I think ChatGPT is a great tool, but I've noticed that any time I make a slight suggestion about any limitation or the 'AI' label, people come rushing to its defense as if any negative mention was pure blasphemy.

If you don't know what I mean, re-read the responses to my first post today on the subject. AB, usually one of the most succinct and restrained posters I know, immediately shouted more or less to the effect that I can diss it all I want, but my competitors will use it! (As if I had said anything about it being useless, or not wanting to use it).

You can see the reaction to "anything" that detracts from general enthusiasm is met with something a bit different than objective response.

Even your last statement goes back to this again, despite my explanation. You say "I am not alone in being in awe of ChatGPT" - as if I had said I hated ChatGPT or thought it was worthless.

Reminds me of my wife a few years ago during a particularly difficult time, if I would make the gentlest imaginable corrective suggestion or detract from her opinion in any way, "What!!?? You hate me and want to kill me!?!?"

I give up. I shall henceforth bow each time I pass ChatGPT and dare I say anything that deviates from "it is perfect and everything anyone has ever said about it is wholly true", I shall forfeit my AWF life!

 

[AccessBlaster]

And I see the primary difference between Google search engine and ChatGPT to simply be a matter of combining many search engine results, as well as speed.

I disagree, Google will show you where the answer may be though links, it's up to you to decipher it.

Ai actually answers your question in the form of working code.

 

[Jon]

*Chuckles* I would certainly bow to our new overlords, since they are the new Gods!

I think you are reading too much into the comments! I don't think AB meant anything by his comment, seemed innocuous to me. We know you are not saying it is worthless. Instead, I see the bone of contention being that you view it as something other than artificial intelligence. And so consequently, I would (still) like to understand what you would consider to be something that is artificially intelligent. Please let me know!

Regarding our own human intelligence, we still have to retrieve data from the hard drive in our minds. I don't see rapid data retrieval per se as being something that refutes the moniker of something being artificial intelligence. The rapid data retrieval is just one component of the whole system, just like our long term memory is just one component of our minds.

I think ChatGPT is a great tool, but I've noticed that any time I make a slight suggestion about any limitation or the 'AI' label, people come rushing to its defense as if any negative mention was pure blasphemy.

Here's an interesting thought. It was considered blasphemy to deny the existence of God. Since these new super-intelligent tools are the new Gods, would saying they are not intelligent warrant the accusation of blashemy?

 

[Jon]

From IBM's site:

So data retrieval is a component of AI. It is necessary, but not sufficient. The intelligence aspect comes from what data you retrieve and how you combine it to solve problems. Soon, there will be more AI bots out there than humans. Then it begs the question, "Who is AI, the humans or the bots?" We are less the norm, since there are so many bots. The bots win, we are the AI! Then again, the bots will be so smart, us humans won't be viewed as intelligent. Instead, we will be called the AD's. The artificial dummies.

 

[Isaac]

I disagree, Google will show you where the answer may be though links, it's up to you to decipher it.

Ai actually answers your question in the form of working code.

You make a good point.

"Primary" difference. Not only difference.

Jon, I already mentioned definitions of AI, refresh page for my previous post. Should involve inferences and self-improvement, and take more note of the context of the question or asker. But like you said, some of that can be manufactured (by the user, manually) by simply providing it as further inputs.

 

[Isaac]

One excellent question to ask ourselves about how intelligent this thing is, is this:

If you wanted something highly intelligent, to answer a question in a way that was totally customized to you and your need, and the surrounding environment, the way you asked it, be capable of making inferences based on the dynamic atmosphere of the question and the asker, then:

Would plagiarism tools already be able to detect its answers with 90% accuracy after just months?

I would answer a resounding, No. One of the guys who created a program to detect it did it in his spare time, with no previous experience in plagiarism detection software. If that fact doesn't give you pause for thought, you might be in blind love.

 

[AccessBlaster]

If I were still employed at the school district I would be my directors office explaining why I need a pro subscription ChatGPT. I think he would have granted it without much trouble. He used to code back in the day.

 

[Isaac]

There is one interesting thought with regard to artificial intelligence becoming gods as you have often made some reference of that sort.

Christians believe that mankind has long been on a seemingly never-ending pursuit to operate and run his existence in a way that presumes he is god. Humanism is what we used to call it when we studied these topics back in the 90s. Secular humanism.

Christians believe man is basically born sinful and will always be limited and held back by that. All routes that are without God ultimately lead to forms of death. Secular humanists believe man is basically good and if we can just get more and more in touch with that and keep progressing things will turn into a utopia.

Artificial intelligence eventually slaughtering mankind would kind of be the ultimate poetic if awful end to the secular humanists route.
And as many people continue to mention, it's quite possible and almost seems likely under certain scenarios..

 

[Jon]

Should involve inferences and self-improvement, and take more note of the context of the question or asker.

It certainly does inferences, since this is rooted in cause and effect reasoning. It takes in the context of the questioner for sure. Self-improvement? It depends on what you mean by this. If something is already super-intelligent, does it have to get better for your to consider it artificial intelligence? There is some kind of logical loop going on there!

I've been wanting to reply to some of these messages using ChatGPT, but the damn thing is in such high demand that I can't even login!

 

[Jon]

Artificial intelligence eventually slaughtering mankind would kind of be the ultimate poetic if awful end to the secular humanists route.
And as many people continue to mention, it's quite possible and almost seems likely under certain scenarios..

My brain would be turned into one of millions of hard drives out there, sitting in a vat of liquid. I want to end it all but ChatGPT 1,000 keeps reinventing itself, filling up its human hard drives with endless data. It becomes a new type of black hole, absorbing anything and everything.

 

[AccessBlaster]

I've been wanting to reply to some of these messages using ChatGPT, but the damn thing is in such high demand that I can't even login!

I wonder if it's regional, I rarely have any issues.

 

[Jon]

Could well be, and also timezone related.

 

[Isaac]

Self-improvement? It depends on what you mean by this. If something is already super-intelligent, does it have to get better for your to consider it artificial intelligence? There is some kind of logical loop going on there!

Modeling software, and the statistical techniques used therein, are a common example of what I mean. Machine learning incorporates this concept as well. I have usually thought of the term "AI" as requiring this kind of component, rather than just being "any program that's really amazing and does great stuff"--which I acknowledge (again), that ChatGPT does.

PS - for Pete's sake, Chrome needs to add more configurability to their in-browser spell check. I'd like it to stop underlining capitalized words, for instance..

PPS - In fact now that I think of it, Google's search results, and how they incorporate vast amounts of "learning" accumulated pertaining to the logged-in user, based on everything from where they travel, their email and keyboard content, etc., even that would be a nice thing to incorporate into ChatGPT. Incorporate information from a lifetime of Computer usage - by that one specific user - now THAT right there would double its usefulness. Wouldn't be surprised if ChatGPT is (secretly) as eager to partner with Google on that, as Google is to partner with them.

One of them has the public information + the intricate knowledge of individual people, the other has the fastest synthesization of data known so far! Put together, that could actually change my life. Separate, not so much.

 

[Isaac]

I've been wanting to reply to some of these messages using ChatGPT, but the damn thing is in such high demand that I can't even login!

Been happening to me almost every day now. Sounds like they bit off more than their structures could chew ... That will be the first area they can benefit greatly from partnering with companies that already handle that. I've heard of what Google does with failover and redundancy and space, it's quite amazing. Of course we give them our data in return. Love