It should be noted that the current state of affairs is likely going to see dramatic improvements in the near future. If ChatGPT ever does incorporate truly "computed responses", such as what Stephen Wolfram has recently suggested doing, that could well be very soon indeed.

Would be very careful with his predictions.
https://www.scientificamerican.com/...ticize-stephen-wolframs-theory-of-everything/

 

Would be very careful with his predictions.
https://www.scientificamerican.com/...ticize-stephen-wolframs-theory-of-everything/

The central premise of the article seems to be the way in which he introduced his theory, rather than pointing out actual errors in his formulation.

Considering the huge accomplishments in computational mathematics, Wolfram still stands as one the true greats. Certainly not someone I would merely brush aside, out of some lack of understanding...

 

The central premise of the article seems to be the way in which he introduced his theory, rather than pointing out actual errors in his formulation.

Considering the huge accomplishments in computational mathematics, Wolfram still stands as one the true greats. Certainly not someone I would merely brush aside, out of some lack of understanding...

It's not a lack of understanding on my part as the guy (who has made huge accomplishments in computational mathematics) has made plenty of IMO fringe predictions before. The problem with ChatGPT and other language models is not computational mathematics. It's a fundamental issue of needing human intelligence generated inputs for it's statistically generated outputs based on queries. It's the classic GIGO with no understanding of the inputs.

 

It's not a lack of understanding on my part as the guy (who has made huge accomplishments in computational mathematics) has made plenty of IMO fringe predictions before. The problem with ChatGPT and other language models is not computational mathematics. It's a fundamental issue of needing human intelligence generated inputs for it's statistically generated outputs based on queries. It's the classic GIGO with no understanding of the inputs.

Did you even read the article? He gave some very convincing examples of how the random Markov chains of ChatGPT can very much right now be coupled with his computational engine, to the effect of getting pretty solid answers out of it. (Or at least as much as one might, say, "trust" a given Wikipedia article.) ChatGPT provides an excellent natural-language interface. Reliable computation is the one thing missing in the whole equation.

 

Did you even read the article? He gave some very convincing examples of how the random Markov chains of ChatGPT can very much right now be coupled with his computational engine, to the effect of getting pretty solid answers out of it. (Or at least as much as one might, say, "trust" a given Wikipedia article.) ChatGPT provides an excellent natural-language interface. Reliable computation is the one thing missing in the whole equation.

Sorry but ChatGPT has no concept of the right and wrong answer so there is no trust factor. Coupling an excellent computational engine like Wolfram|Alpha will just result in mathematical AI hallucinations and lies generated from the ChatGPT natural-language interface.

 

https://www.nature.com/articles/d41586-023-00107-z

The editors-in-chief of Nature and Science told Nature’s news team that ChatGPT doesn’t meet the standard for authorship. “An attribution of authorship carries with it accountability for the work, which cannot be effectively applied to LLMs,” says Magdalena Skipper, editor-in-chief of Nature in London. Authors using LLMs in any way while developing a paper should document their use in the methods or acknowledgements sections, if appropriate, she says.

“We would not allow AI to be listed as an author on a paper we published, and use of AI-generated text without proper citation could be considered plagiarism,” says Holden Thorp, editor-in-chief of the Science family of journals in Washington DC.

Zhavoronkov says that when he tried to get ChatGPT to write papers more technical than the perspective he published, it failed. “It does very often return the statements that are not necessarily true, and if you ask it several times the same question, it will give you different answers,” he says. “So I will definitely be worried about the misuse of the system in academia, because now, people without domain expertise would be able to try and write scientific papers.”

 

https://spectrum.ieee.org/rodney-brooks-ai

An Inconvenient Truth About AI
AI won't surpass human intelligence anytime soon

Sometimes we are in the loop even when the consequences of failure aren't dire. AI systems power the speech and language understanding of our smart speakers and the entertainment and navigation systems in our cars. We, the consumers, soon adapt our language to each such AI agent, quickly learning what they can and can't understand, in much the same way as we might with our children and elderly parents. The AI agents are cleverly designed to give us just enough feedback on what they've heard us say without getting too tedious, while letting us know about anything important that may need to be corrected. Here, we, the users, are the people in the loop. The ghost in the machine, if you will.

Ask not what your AI system can do for you, but instead what it has tricked you into doing for it.

 

https://www.theverge.com/23560328/openai-gpt-4-rumor-release-date-sam-altman-interview
OpenAI CEO Sam Altman on GPT-4: ‘people are begging to be disappointed and they will be’

We don’t have an actual AGI and that’s sort of what’s expected of us.”
(AGI here refers to “artificial general intelligence” — shorthand for an AI system with at least human-equivalent capabilities across many domains.)

When asked about one viral (and factually incorrect) chart that purportedly compares the number of parameters in GPT-3 (175 billion) to GPT-4 (100 trillion), Altman called it "complete bulls$it."

"The GPT-4 rumor mill is a ridiculous thing. I don't know where it all comes from," said the OpenAI CEO. "People are begging to be disappointed and they will be. The hype is just like... We don't have an actual AGI and that's sort of what's expected of us."
Asked about how far we are from developing AGI, Altman replied "The closer we get, the harder time I have answering. Because I think it's going to be much blurrier and much more of a gradual transition than people think."

And Altman also addressed predictions that ChatGPT will kill Google. "I think whenever someone talks about a technology being the end of some other giant company, it's usually wrong. I think people forget they get to make a countermove here, and they're like pretty smart, pretty competent. I do think there's a change for search that will probably come at some point — but not as dramatically as people think in the short term."

 

https://www.scientificamerican.com/article/research-summaries-written-by-ai-fool-scientists/
Research Summaries Written by AI Fool Scientists


More believable lies. Nonsense that looks good, but isn't good.

But we have to keep in mind that the devil is in the details, not in the abstract. Research abstracts are often, by their very nature, difficult for experts to discern whether there is anything there or not. The whole point is that the paper that is being abstracted is adding some new knowledge to the general body of knowledge. It's the paper, not the abstract, that has to stand up to scrutiny. Even beyond that, the paper is not the end-all, be-all -- the paper needs to be written so as to allow independent confirmation of the claimed results.

What will be exciting is when one of these AI engines can draw upon the data at its disposal and, however it happens to do it, generate something NEW that is not only verifiable based on what it produces, but that is then successfully verified.

 

What will be exciting is when one of these AI engines can draw upon the data at its disposal and, however it happens to do it, generate something NEW that is not only verifiable based on what it produces, but that is then successfully verified.

And then said feat must be repeatable to make sure the first one was not a fluke. That is, a statistically significant sequence of successful predictions/discoveries must happen in order to declare AI as being "real" in a sense of being truly intelligent. Although not necessarily conscious.

 

And then said feat must be repeatable to make sure the first one was not a fluke. That is, a statistically significant sequence of successful predictions/discoveries must happen in order to declare AI as being "real" in a sense of being truly intelligent. Although not necessarily conscious.

Sure. Also, I suspect that IF it ever happens (and, personally, I think that's a pretty big if), it will be interesting to see if it is really something novel and new, or if it is "merely" pulling strings of knowledge that are already out there together and up to the surface. I suspect it will be the latter, but that would still be remarkable and extremely useful.

 

So far I am very impressed with ChatGPT, it's often dead wrong on the facts, but as a productivity tool, I am already hooked.

It can only improve from here.

 

So far I am very impressed with ChatGPT, it's often dead wrong on the facts, but as a productivity tool, I am already hooked.

It can only improve from here.

And from a big-data science achievement, it's right on the money. Volume, variety, and velocity.

The first thing that hit me when I heard about the "three Vs" of big data was that there was no mention of validity.

 

So far I am very impressed with ChatGPT, it's often dead wrong on the facts, but as a productivity tool, I am already hooked.

It can only improve from here.

Or eventually be another useful ability humans can manipulate like most machines capabilities but a dead-end to an actual AGI.

It's quite seductive as ChatGPT gives the illusion that it is "thinking" due to its ability to both parse and compose language based on word relationship. Always remember, it is a trick designed by very smart people that produces wrong answers just as easily as correct ones.

 

But we have to keep in mind that the devil is in the details, not in the abstract. Research abstracts are often, by their very nature, difficult for experts to discern whether there is anything there or not. The whole point is that the paper that is being abstracted is adding some new knowledge to the general body of knowledge. It's the paper, not the abstract, that has to stand up to scrutiny. Even beyond that, the paper is not the end-all, be-all -- the paper needs to be written so as to allow independent confirmation of the claimed results.

What will be exciting is when one of these AI engines can draw upon the data at its disposal and, however it happens to do it, generate something NEW that is not only verifiable based on what it produces, but that is then successfully verified.

The types of applications I see for ChatCPT are not NEW discoveries. It's analyzing massive amounts of data to produce abstracts and summaries that can only bee seen as flawed by high level human expertise, not other machines.

https://thehill.com/policy/technolo...-wharton-business-school-test-research-paper/
ChatGPT passes Wharton Business School test

Research from Wharton professor Christian Terwiesch found that the AI system “has shown a remarkable ability to automate some of the skills of highly compensated knowledge workers in general and specifically the knowledge workers in the jobs held by MBA graduates including analysts, managers, and consultants.”

On the final exam of Operations Management, a core course in the Wharton MBA program, ChatGPT did “an amazing job” and gave answers that were correct and “excellent” in their explanations.

“ChatGPT3 is remarkably good at modifying its answers in response to human hints. In other words, in the instances where it initially failed to match the problem with the right solution method, Chat GPT3 was able to correct itself after receiving an appropriate hint from a human expert. Considering this performance, Chat GPT3 would have received a B to B- grade on the exam,” the research concluded.

The human expert enabled that grade by iteration and elimination of incorrect and bad answers. Without the human intelligence in the loop as the Oracle, the system easily produces convincing lies.

 

https://www.zdnet.com/article/chatg...-revolutionary-says-metas-chief-ai-scientist/
ChatGPT is 'not particularly innovative,' and 'nothing revolutionary', says Meta's chief AI scientist

Why hasn't the public seen programs like ChatGPT from Meta or from Google? "The answer is, Google and Meta both have a lot to lose by putting out systems that make stuff up," says Meta's chief AI scientist, Yann LeCun.

+1
Buy that man a jelly donut. LeCun has lived through a couple iterations of somebody putting together an impressive AI demo, promising the moon, and discrediting the entire field.

 

https://slate.com/technology/2023/01/ai-chatgpt-scientific-literature-peer-review.html
A.I. Like ChatGPT Is Revealing the Insidious Disease at the Heart of Our Scientific Process

What ChatGPT is doing, like SCIgen before it, is generating BS that should (at least at this point) fail a serious peer review. Sure, ChatGPT and its A.I. siblings are orders of magnitude more sophisticated than SCIgen, but if you cut through the verbiage, much of what these programs produce is wrong, even nonsensical. They make up facts and generate fake sources to back up their lies. A proper peer review, which checks references and vets facts, should catch the problem.

Take, for example, one of the peer-reviewed papers Nature described—“Can GPT-3 write an academic paper on itself, with minimal human input?” Posted on a preprint server in July, the article was reportedly rejected by one journal and then accepted to another with revisions. While we don’t yet know the nature of those revisions, the paper that the authors have made public is full of errors that a competent peer review should catch. The references are a mess: Two seem to be referencing the same online article in different ways—an article that doesn’t appear to exist. And if it did exist, it couldn’t contain the claimed information about the GPT-3 bot, as the reference dates from four years before it first premiered. Similarly, another reference seems made up out of whole cloth. Two of the referenced works do exist, but a cursory check shows that the contexts of those references are entirely bogus: They don’t say what the peer-reviewed paper says they say. (The researchers submitting the paper did, happily, notice that the references were nonsense.) More important: The A.I.-generated text was anodyne and information-free; it added nothing to anyone’s understanding of anything. The bot is constitutionally unable to advance knowledge in the field because it’s fundamentally a recycling engine, regurgitating what it’s been trained with in a controlled manner. That’s not the way to advance knowledge. By definition, it doesn’t belong in a peer review journal.

In other words, if the A.I. apocalypse is here, it’s not so much a Terminator as it is an Idiocracy.

 

Sure, ChatGPT and its A.I. siblings are orders of magnitude more sophisticated than SCIgen, but if you cut through the verbiage, much of what these programs produce is wrong, even nonsensical. They make up facts and generate fake sources to back up their lies. A proper peer review, which checks references and vets facts, should catch the problem.

We've been living with that for what, 7 years? Never knew it had a different scientific name.

 

Related to the subject being discussed:

AI-powered investment fund beats the market by almost 100%

​

Instead of simply relying on standard market data as available from the likes of Bloomberg or S&P, AIEQ also looks through unstructured data such as earnings calls, keyword data, and tweets, Chris Natividad, Chief Investment Officer at Equbot told ETF.com

I predict a huge economic bubble followed by an unprecedented gargantuan burst ...

 

So far I am very impressed with ChatGPT, it's often dead wrong on the facts, but as a productivity tool, I am already hooked.

It can only improve from here.

Or not. AI often has sexy promises which turn to be false beyond a very limited set of circumstances.

Back in the late 60s and early 70s, I wrote a paper on natural language AI. At first, I provided several examples of its intelligence. Then, illustrated several flaws. And concluded by tricking it into “chatting” in pure garbage. “The glenidh is mawsken ferkloinish.”