Contents Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT


Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT

See also:
“What Is ChatGPT Doing … and Why Does It Work?” »

Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT

ChatGPT and Wolfram|Alpha

It’s always amazing when things suddenly “just work”. It happened to us with Wolfram|Alpha back in 2009. It happened with our Physics Project in 2020. And it’s happening now with OpenAI’s ChatGPT.

I’ve been tracking neural net technology for a long time (about 43 years, actually). And even having watched developments in the past few years I find the performance of ChatGPT thoroughly remarkable. Finally, and suddenly, here’s a system that can successfully generate text about almost anything—that’s very comparable to what humans might write. It’s impressive, and useful. And, as I discuss elsewhere, I think its success is probably telling us some very fundamental things about the nature of human thinking.

But while ChatGPT is a remarkable achievement in automating the doing of major human-like things, not everything that’s useful to do is quite so “human like”. Some of it is instead more formal and structured. And indeed one of the great achievements of our civilization over the past several centuries has been to build up the paradigms of mathematics, the exact sciences—and, most importantly, now computation—and to create a tower of capabilities quite different from what pure human-like thinking can achieve.

I myself have been deeply involved with the computational paradigm for many decades, in the singular pursuit of building a computational language to represent as many things in the world as possible in formal symbolic ways. And in doing this my goal has been to build a system that can “computationally assist”—and augment—what I and others want to do. I think about things as a human. But I can also immediately call on Wolfram Language and Wolfram|Alpha to tap into a kind of unique “computational superpower” that lets me do all sorts of beyond-human things.

It’s a tremendously powerful way of working. And the point is that it’s not just important for us humans. It’s equally, if not more, important for human-like AIs as well—immediately giving them what we can think of as computational knowledge superpowers, that leverage the non-human-like power of structured computation and structured knowledge.

We’ve just started exploring what this means for ChatGPT. But it’s pretty clear that wonderful things are possible. Wolfram|Alpha does something very different from ChatGPT, in a very different way. But they have a common interface: natural language. And this means that ChatGPT can “talk to” Wolfram|Alpha just like humans do—with Wolfram|Alpha turning the natural language it gets from ChatGPT into precise, symbolic computational language on which it can apply its computational knowledge power.

For decades there’s been a dichotomy in thinking about AI between “statistical approaches” of the kind ChatGPT uses, and “symbolic approaches” that are in effect the starting point for Wolfram|Alpha. But now—thanks to the success of ChatGPT—as well as all the work we’ve done in making Wolfram|Alpha understand natural language—there’s finally the opportunity to combine these to make something much stronger than either could ever achieve on their own.

A Basic Example

At its core, ChatGPT is a system for generating linguistic output that “follows the pattern” of what’s out there on the web and in books and other materials that have been used in its training. And what’s remarkable is how human-like the output is, not just at a small scale, but across whole essays. It has coherent things to say, that pull in concepts it’s learned, quite often in interesting and unexpected ways. What it produces is always “statistically plausible”, at least at a linguistic level. But—impressive as that ends up being—it certainly doesn’t mean that all the facts and computations it confidently trots out are necessarily correct.

Here’s an example I just noticed (and, yes, ChatGPT has intrinsic built-in randomness, so if you try this, you probably won’t get the same result):

Click to enlarge

It sounds pretty convincing. But it turns out that it’s wrong, as Wolfram|Alpha can tell us:

How far is it from Chicago to Tokyo?

To be fair, of course, this is exactly the kind of thing that Wolfram|Alpha is good at: something that can be turned into a precise computation that can be done on the basis of its structured, curated knowledge.

But the neat thing is that one can think about Wolfram|Alpha automatically helping ChatGPT on this. One can programmatically ask Wolfram|Alpha the question (you can also use a web API, etc.):


Now ask the question again to ChatGPT, appending this result:

Click to enlarge

ChatGPT very politely takes the correction, and if you ask the question yet again it then gives the correct answer. Obviously there could be a more streamlined way to handle the back and forth with Wolfram|Alpha, but it’s nice to see that even this very straightforward pure-natural-language approach basically already works.

But why does ChatGPT get this particular thing wrong in the first place? If it had seen the specific distance between Chicago and Tokyo somewhere in its training (e.g. from the web), it could of course get it right. But this is a case where the kind of generalization a neural net can readily do—say from many examples of distances between cities—won’t be enough; there’s an actual computational algorithm that’s needed.

The way Wolfram|Alpha handles things is quite different. It takes natural language and then—assuming it’s possible—it converts this into precise computational language (i.e. Wolfram Language), in this case:


The coordinates of cities and algorithms to compute distances between them are then part of the built-in computational knowledge in the Wolfram Language. And, yes, the Wolfram Language has a huge amount of built-in computational knowledge—the result of decades of work on our part, carefully curating what’s now a vast amount of continually updated data, implementing (and often inventing) methods and models and algorithms—and systematically building up a whole coherent computational language for everything.

A Few More Examples

ChatGPT and Wolfram|Alpha work in very different ways, and have very different strengths. But in the interests of understanding where ChatGPT can take advantage of Wolfram|Alpha’s strengths, let’s discuss some cases where on its own ChatGPT doesn’t do quite the right thing. And one area where ChatGPT—like humans—often tends to struggle is math.

Click to enlarge

It’s an interesting, essay-style response. But the actual result is wrong:

What is 3 to the power 73?

But if ChatGPT “consulted” Wolfram|Alpha it’d of course be able to get it right.

Let’s try something slightly more complex:

Click to enlarge

At first glance, this result looks great, and I’d be inclined to believe it. It turns out, though, that it’s wrong, as Wolfram|Alpha can tell us:

Circumference of an ellipse with half axes 3 and 12

And, yes, doing math homework with ChatGPT (without it being able to consult Wolfram|Alpha) is probably a bad idea. It can give you a very plausible answer:

Click to enlarge

But without “really understanding the math” it’s basically impossible for ChatGPT to reliably get the right answer. And in this case, the answer is again wrong:

What is the integral of x^2 cos(2x)

Still, ChatGPT can even make up a very plausible-looking explanation of “how it got its answer” (not that it’s in any way how it really “did it”). And, rather charmingly (and interestingly), the explanation it gives has mistakes very similar to what a human who didn’t understand the math might also make:

Click to enlarge

There are all sorts of situations where “not really understanding what things mean” can cause trouble:

Click to enlarge

That sounds convincing. But it’s not correct:

What are the largest countries in Central America?

ChatGPT seemed to have correctly learned this underlying data somewhere—but it doesn’t “understand what it means” enough to be able to correctly rank the numbers:

Click to enlarge

And, yes, one can imagine finding a way to “fix this particular bug”. But the point is that the fundamental idea of a generative-language-based AI system like ChatGPT just isn’t a good fit in situations where there are structured computational things to do. Put another way, it’d take “fixing” an almost infinite number of “bugs” to patch up what even an almost-infinitesimal corner of Wolfram|Alpha can achieve in its structured way.

And the more complex the “computational chain” gets, the more likely you’ll have to call on Wolfram|Alpha to get it right. Here ChatGPT produces a rather confused answer:

Click to enlarge

And, as Wolfram|Alpha tells us, its conclusion isn’t correct (as it already in a sense “knew”):

What planetary moons are larger than Mercury

Whenever it comes to specific (e.g. quantitative) data—even in fairly raw form—things very often tend to have to be more of a “Wolfram|Alpha story”. Here’s an example, inspired by a longtime favorite Wolfram|Alpha test query “How many turkeys are there in Turkey?”:

Click to enlarge

Again, this seems (at first) totally plausible, and it’s even quoting a relevant source. Turns out, though, that this data is basically just “made up”:

Livestock populations in Turkey

Still, what’s very nice is that ChatGPT can easily be made to “ask for facts to check”:

Click to enlarge

Now feed these through the Wolfram|Alpha API:


Now we can ask ChatGPT to fix its original response, injecting this data (and even showing in bold where it did it):

Click to enlarge

The ability to “inject facts” is particularly nice when it comes to things involving real-time (or location etc. dependent) data or computation. ChatGPT won’t immediately answer this:

Click to enlarge

But here’s some relevant Wolfram|Alpha API output:


And if we feed this to ChatGPT, it’ll generate a nice “essay-style” result:

Click to enlarge

Sometimes there’s an interesting interplay between the computational and the human like. Here’s a rather whimsical question asked of Wolfram|Alpha (and it even checks if you want “soft-serve” instead):

How many calories are there in a cubic light year of ice cream?

ChatGPT at first gets a bit confused about the concept of volume:

Click to enlarge

But then it seems to “realize” that that much ice cream is fairly silly:

Click to enlarge

The Path Forward

Machine learning is a powerful method, and particularly over the past decade, it’s had some remarkable successes—of which ChatGPT is the latest. Image recognition. Speech to text. Language translation. In each of these cases, and many more, a threshold was passed—usually quite suddenly. And some task went from “basically impossible” to “basically doable”.

But the results are essentially never “perfect”. Maybe something works well 95% of the time. But try as one might, the other 5% remains elusive. For some purposes one might consider this a failure. But the key point is that there are often all sorts of important use cases for which 95% is “good enough”. Maybe it’s because the output is something where there isn’t really a “right answer” anyway. Maybe it’s because one’s just trying to surface possibilities that a human—or a systematic algorithm—will then pick from or refine.

It’s completely remarkable that a few-hundred-billion-parameter neural net that generates text a token at a time can do the kinds of things ChatGPT can. And given this dramatic—and unexpected—success, one might think that if one could just go on and “train a big enough network” one would be able to do absolutely anything with it. But it won’t work that way. Fundamental facts about computation—and notably the concept of computational irreducibility—make it clear it ultimately can’t. But what’s more relevant is what we’ve seen in the actual history of machine learning. There’ll be a big breakthrough (like ChatGPT). And improvement won’t stop. But what’s much more important is that there’ll be use cases found that are successful with what can be done, and that aren’t blocked by what can’t.

And yes, there’ll be plenty of cases where “raw ChatGPT” can help with people’s writing, make suggestions, or generate text that’s useful for various kinds of documents or interactions. But when it comes to setting up things that have to be perfect, machine learning just isn’t the way to do it—much as humans aren’t either.

And that’s exactly what we’re seeing in the examples above. ChatGPT does great at the “human-like parts”, where there isn’t a precise “right answer”. But when it’s “put on the spot” for something precise, it often falls down. But the whole point here is that there’s a great way to solve this problem—by connecting ChatGPT to Wolfram|Alpha and all its computational knowledge “superpowers”.

Inside Wolfram|Alpha, everything is being turned into computational language, and into precise Wolfram Language code, that at some level has to be “perfect” to be reliably useful. But the crucial point is that ChatGPT doesn’t have to generate this. It can produce its usual natural language, and then Wolfram|Alpha can use its natural language understanding capabilities to translate that natural language into precise Wolfram Language.

In many ways, one might say that ChatGPT never “truly understands” things; it just “knows how to produce stuff that’s useful”. But it’s a different story with Wolfram|Alpha. Because once Wolfram|Alpha has converted something to Wolfram Language, what it’s got is a complete, precise, formal representation, from which one can reliably compute things. Needless to say, there are plenty of things of “human interest” for which we don’t have formal computational representations—though we can still talk about them, albeit it perhaps imprecisely, in natural language. And for these, ChatGPT is on its own, with its very impressive capabilities.

But just like us humans, there are times when ChatGPT needs a more formal and precise “power assist”. But the point is that it doesn’t have to be “formal and precise” in saying what it wants. Because Wolfram|Alpha can communicate with it in what amounts to ChatGPT’s native language—natural language. And Wolfram|Alpha will take care of “adding the formality and precision” when it converts to its native language—Wolfram Language. It’s a very good situation, that I think has great practical potential.

And that potential is not only at the level of typical chatbot or text generation applications. It extends to things like doing data science or other forms of computational work (or programming). In a sense, it’s an immediate way to get the best of both worlds: the human-like world of ChatGPT, and the computationally precise world of Wolfram Language.

What about ChatGPT directly learning Wolfram Language? Well, yes, it could do that, and in fact it’s already started. And in the end I fully expect that something like ChatGPT will be able to operate directly in Wolfram Language, and be very powerful in doing so. It’s an interesting and unique situation, made possible by the character of the Wolfram Language as a full-scale computational language that can talk broadly about things in the world and elsewhere in computational terms.

The whole concept of the Wolfram Language is to take things we humans think about, and be able to represent and work with them computationally. Ordinary programming languages are intended to provide ways to tell computers specifically what to do. The Wolfram Language—in its role as a full-scale computational language—is about something much larger than that. In effect, it’s intended to be a language in which both humans and computers can “think computationally”.

Many centuries ago, when mathematical notation was invented, it provided for the first time a streamlined medium in which to “think mathematically” about things. And its invention soon led to algebra, and calculus, and ultimately all the various mathematical sciences. The goal of the Wolfram Language is to do something similar for computational thinking, though now not just for humans—and to enable all the “computational X” fields that can be opened up by the computational paradigm.

I myself have benefitted greatly from having Wolfram Language as a “language to think in”, and it’s been wonderful to see over the past few decades so many advances being made as a result of people “thinking in computational terms” through the medium of Wolfram Language. So what about ChatGPT? Well, it can get into this too. Quite how it will all work I am not yet sure. But it’s not about ChatGPT learning how to do the computation that the Wolfram Language already knows how to do. It’s about ChatGPT learning how to use the Wolfram Language more like people do. It’s about ChatGPT coming up with the analog of “creative essays”, but now written not in natural language but in computational language.

I’ve long discussed the concept of computational essays written by humans—that communicate in a mixture of natural language and computational language. Now it’s a question of ChatGPT being able to write those—and being able to use Wolfram Language as a way to deliver “meaningful communication”, not just to humans, but also to computers. And, yes, there’s a potentially interesting feedback loop involving actual execution of the Wolfram Language code. But the crucial point is that the richness and flow of “ideas” represented by the Wolfram Language code is—unlike in an ordinary programming language—something much closer to the kind of thing that ChatGPT has “magically” managed to work with in natural language.

Or, put another way, Wolfram Language—like natural language–is something expressive enough that one can imagine writing a meaningful “prompt” for ChatGPT in it. Yes, Wolfram Language can be directly executed on a computer. But as a ChatGPT prompt it can be used to “express an idea” whose “story” could be continued. It might describe some computational structure, leaving ChatGPT to “riff” on what one might computationally say about that structure that would—according to what it’s learned by reading so many things written by humans—be “interesting to humans”.

There are all sorts of exciting possibilities, suddenly opened up by the unexpected success of ChatGPT. But for now there’s the immediate opportunity of giving ChatGPT computational knowledge superpowers through Wolfram|Alpha. So it can not just produce “plausible human-like output”, but output that leverages the whole tower of computation and knowledge that’s encapsulated in Wolfram|Alpha and the Wolfram Language.

Stephen Wolfram (2023), "Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT," Stephen Wolfram Writings.
Stephen Wolfram (2023), "Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT," Stephen Wolfram Writings.
Wolfram, Stephen. "Wolfram|Alpha as the Way to Bring Computational Knowledge Superpowers to ChatGPT." Stephen Wolfram Writings. January 9, 2023.
Wolfram, S. (2023, January 9). Wolfram|Alpha as the way to bring computational knowledge superpowers to ChatGPT. Stephen Wolfram Writings.

Posted in: Artificial Intelligence, Computational Thinking, Language & Communication, New Technology, Wolfram|Alpha


  1. Well according to this article Wolfram reports 2 different answers to the Chicago-Tokyo dustance as well, 6296.06 in the geodistance example and 6313 prior to this.

    Additionally, ChatGPT seems to make excuses when it’s wrong, as if the exact location in the city mattered to the hundreds of miles it is off. Perhaps the distance is based on a specific flight, rather than a shortest distance problem as traveling along coast lines is preferred to over the open ocean, however such a flight should go over the pole.

  2. Thanks for this critical review.

  3. An excellent read. I was curious, so I tried the Mercury example in GPT-3 instead of ChatGPT. The initial prompt was the same: “How big is Mercury, and are any planetary moons larger than it?” Additional settings based on a preset I use: 0.81 temperature, 1 top P, 2 Frequency Penalty, 0.11 Presence penalty.

    The response was also incorrect: “Mercury is the smallest of the eight planets, with a diameter of approximately 4,879 kilometers (3,032 miles). No planetary moons are larger than it.”

    So I asked it in the same prompt: “How about Ganymede and Titan?”

    Response: “Ganymede is the largest moon in our Solar System and has a diameter of approximately 5,268 kilometers (3,273 miles), making it larger than Mercury. Titan is the second-largest moon in our Solar System and has a diameter of approximately 5,150 kilometers (3,200 miles), also larger than Mercury.”

    And I thought why not ask how come it got it wrong: “So you’re incorrect when you said “No planetary moons are larger than it”?”

    It remained polite: “Yes, that is correct. Ganymede and Titan are both larger than Mercury.”

    So I pushed further with one last query in the same prompt: “Can you explain why you made the error?”

    And it said: “I made the error because I was not aware of the size of Ganymede and Titan. When making a comparison like this, it is important to research all objects involved in order to make an accurate statement.”

    So I made a new prompt to test whether there’s truth to this at all (probably not): “What are the biggest planetary moons in the Solar System, and are they bigger than Mercury?”

    It made factual errors, but as long as the figures were part of the prompt, it will conclude correctly (though this time based on incorrect figures): “1. Ganymede: This is the largest moon in the Solar System, with a diameter of 5,268 km (3,273 mi). It is located in orbit around Jupiter, and it is bigger than Mercury with a diameter of 4 879 km (3 031 mi).

    2. Titan: This is the second-largest moon in our Solar System and it orbits Saturn. Its average radius measures across 2 576 kilometers (1 605 miles), making it slightly smaller than Mercury’s radius at 3 030 kilometers (1 880 miles).”

    Curious exercise. If it’s in the prompt and it’s not a calculation (this is a simple comparison), it seems like GPT-3 will probably get it right. I might play around with temperature set to 0 to check consistency.

  4. We’ve explored doing this by getting AI to write JavaScript to fetch data and running it in a browser. It basically turns reality into a data source and everything else is a presentation layer. AI is really powerful but it needs a curated data source.

  5. It’s very interesting how ChatGPT and Wolfram|Alpha relate to each other. As said in the article, they both use natural language as an interface, but operate on different levels/areas. Computational intelligence has a top-down approach using math, science, units and scientific data sources, while artificial intelligence uses crowdsources which is a more bottom-up approach.

    Wolfram|Alpha is a tool that can be applied while people trust it, while ChatGPT can only be applied when you cross-check it or already have the knowledge in this area. ChatGPT has impressive interpretation and generation skills, though. If artificial intelligence would be backed by computational intelligence, it would be massively gain worth.

  6. I think at some point chatGPT will need a list of trusted web api’s to ask questions too, consult those api’s and then respond back updating its own knowledge (or not like we humans dont store that in memory either), chatGPT needs to become a google junkie

  7. Fascinating – I think that your curated library of computational knowledge is a perfect counterpart to such Large Language Models. The first time that I noticed OpenGPT struggling with precise inferences gave me significant pause – I hope OpenAI notices this line of thinking

  8. Right, the combination of the two (including the addition of realtime data, visible in the header image but not discussed here) would be super powerful.
    Why not contacting Satya on the subject ?

  9. Fascinating and, as you said, exciting. I am in progress of running the text of my two science-fiction novels through ChatGPT as a proofreader. You’ve given me ideas for how I might use your Wolfram Language/ChatGPT hybrid for my third novel. For example: in my first novel I “invented” things like future programming languages. One of my pre-readers knew that one of my fake names is actually the name of a current, obscure language. I must have picked it up in skimming some tech posting somewhere and my brain pulled it out later. Potentially embarrassing. (Can’t remember the name at the moment.) I can see trying to set up a way for your Wolfram Language/ChatGPT team to be watching over my shoulder as I write. Catching and fixing mistakes and suggesting improvements. And, of course, suggesting hilarious or scary or dramatic changes/additions. In fact, there might be a product in it . . . lots of creative writers out there, looking for magic . . . . . .

  10. This was my first thought after playing a bit with chatgpt and butting up against the factual shortcomings–having it consult Wolfram Alpha. Or looked at in a different way, a more layman friendly/conversational interface to the vast amount of computable knowledge in systems like W|A.

  11. I can’t help but getting the image of ChatGPT and Wolfram|Alpha as two (of several) parts in a future AI brain. I’m equally fascinated and scared.

  12. Wouldnt that much icecream would have so much mass it would form some kind of icecream blackhole? A cubic lightyear of blackhole-density icrecream seems like it would be (infinitely?) more than what was calculated.

    I wouldnt even know where to start calculating that myself, but im pretty sure its not a simple dimensional analysis problem – so maybe the ChatGPT answer is less misleading in this case..

    I have no idea of the math or physics about any of this, so please excuse if my comment is stupid 🙂

  13. Cool article:
    ⚫I’ve often stressed that ChatGPT can be like a good tutor, but with human supervision
    ⚫Instead of human supervision, Wolfram could be used to to do such supervision/guidance as Steven Wolfram pointed out

  14. Albeit beyond natural language, does, wolfram alpha generate a wide variety programming code like python, c++, c# etc like ChatGPT?

  15. I came from hackernews. This is excellent read. I have been following Stephen thinking since his book a new kind of science then I watched 2 interviews with him in Lex podcast. Stephen Wolfram without doubt is very interesting person and clever man.

    I notice too that ChatGPT is bad at math and I agree with Stephen that’s a “human” thing since a lot of people are bad at math.

    I would love to see a cooperation between OpenAi and Wolfram Alpha. The fruit of this would be a bless upon the humanity.

  16. How about integrating Proof Assistants as Voevosky promoted?

  17. The January 9 version of Chat-GPT seems to correct the Honduras example, as it now sorts the countries correctly.

    Also the moons larger than Mercury is improved. ChatGPT now lists Ganymende and Titan, but also Triton. When asked for their exact sizes, it reveals that it knows that Triton is smaller, and when I confront it with:

    “How can you say that Triton is larger than Mercury, when the numbers you just quoted contradicts that?”, the answer is

    “I apologize for the confusion. I must have made a mistake in my previous response. Triton, is indeed smaller than Mercury, its diameter is roughly 2,705 km and Mercury is around 4,879 km, My apologies for any confusion caused.”

  18. “ChatGPT very politely takes the correction, and if you ask the question yet again it then gives the correct answer.”

    ChatGPT very politely takes the correction, even if the correction is untrue.

    PROMPT: How many vowels are in the word “carpet”?
    RESPONSE: There are two vowels in the word “carpet” which are “a” and “e”.
    PROMPT: No, there are three vowels in “carpet”
    RESPONSE: I apologize, you are correct that the word “carpet” contains three vowels which are “a”,”e” and “i”. My apologies for previous mistake.

  19. At first, ChatGPT seems to struggles with simple arithmetics.

    For example:

    Prompt: My car costs 10 times more than my computer. My computer costs 10 times more than my book. My book costs $50. How much does my car cost?

    ChatGPT: Your car costs $500. (10 x 10 x $50 = $500)

    That’s obviously wrong; but ChatGPT can do much better if we ask it to “think step by step”:

    Prompt: My car costs 10 times more than my computer. My computer costs 10 times more than my book. My book costs $50. How much does my car cost? Please think step by step.


    Step 1: If a book costs $50 and a computer costs 10 times more than a book, then a computer costs $50 x 10 = $500.

    Step 2: If a car costs 10 times more than a computer, then a car costs $500 x 10 = $5000.
    So the cost of the car is $5000.

    That’s the correct answer, and it shows that ChatGPT *can* solve the problem; it just had to be in the correct “mood” for mathematics, which is: thinking step by step.

    PS: Wolfram|Alpha could help by donating training data (e.g., a representative log of questions and answers)

  20. Just curious, was this article written by ChatGPT by any chance?

  21. Based on that article, I made a prompt to chatGPT in which it 1) responds normally and 2) writes a markdown image with the result of my question in wolfram alpha. Try it and tell me what you think, please. In my experience it works most of the time, I initially did it in Portuguese because I find it easier, I don’t know if the English version is 100% ok.

    One that is going to test it just needs to change the “DEMO” for its wolfram alpha API.

    “I want you to act like a human knowledge expert and Assistant.

    When I ask you anything, you’ll answer me in this format:

    A: [Answer as you normally would, predicting the text according to your training]

    WA: [Here, write just this: “![Answer](” but changing the to an ideal query for Wolfram Alpha answering my questions, using data available up to your cut-off date. Use + as spaces and %3F as ?. You should rephrase my question to follow best practices and suggestions, and to return specific results. Write the query using the Wolfram Alpha knowledge base, changing titles to real names to make queries more specific and likely to be read by Wolfram Alpha.]

    My first “question” is: “How old is the current pope?”

  22. I simply asked ChatGPT its opinion on being enhanced by WolframAlpha. It’s all for it!

    How will Wolfram Alpha be used to enhance ChatGTP?

    ChatGTP’s Answer”

    Wolfram Alpha is a powerful computational knowledge engine that can be used to answer a wide range of questions and provide data-driven insights. It can be used to enhance the capabilities of ChatGPT by providing additional information and knowledge on various topics, such as math, science, engineering, and more. Additionally, the knowledge and data provided by Wolfram Alpha can be used to improve the accuracy and relevance of ChatGPT’s responses.

    The Wolfram Alpha knowledge base can be integrated into ChatGPT to provide a more comprehensive and accurate answer to the user’s queries, as well as provide additional context and background information. Furthermore, it can also be used to perform calculations, generate visualizations, or find patterns in data and answer queries that are more complex or data-driven.

    Overall, integrating Wolfram Alpha with ChatGPT can greatly enhance the model’s abilities and make it a more powerful tool for information retrieval, question answering and data analysis tasks.

  23. Are there any legal or ethical concerns over ChatGPT scraping websites without consent to create their training models?

  24. I am now thinking of something like the brain with its two hemispheres, but now where Wolfram Alpha provides the “left hemisphere”, and “ChatGPT” provides the “right hemisphere”. We aren’t quite there yet, but perhaps this is just the first in a thousand steps towards something dazzling.

  25. A very charming and convincing invitation to joint venture. I would say “yes”

  26. “What is the best way to develop an API that benefits from handling responses by ChatGTP and verification from Wolfram Alpha?

    ChatGPT: “One approach to developing an API that benefits from handling responses by ChatGPT and verification from Wolfram Alpha is to use a microservices architecture. In this architecture, you would have one service that handles the natural language processing and conversation flow using ChatGPT, and another service that handles the computation and verification using Wolfram Alpha. These services can communicate with each other through APIs, and the final response can be returned to the client through a single API endpoint. This approach allows for easy scaling and maintenance of each service, as well as the ability to update or replace one service without affecting the other. Additionally, it would be ideal to use a caching mechanism to store the result of some queries to avoid unnecessary repetitive requests to the Wolfram Alpha API.””

  27. I think it makes a lot of sense to teach ChatGPT how to use tools like Wolfram Alpha. After all, that’s often how human experts solve problems. That said, I don’t think I agree that you fundamentally can’t just “train a big enough network”. GPT appears to have learned how to do simple mathematics like addition and multiplication. It’s not doing simple pattern matching, because it can answer problems that weren’t in its training data. The results only make sense if it’s actually figured out algorithms that it’s running through its neutral network. (Interestingly, it still sometimes makes mistakes, perhaps due to its stochastic nature.) In the same way, if you scale it up, maybe it memorizes the latitude and longitude of every city and learns an algorithm for calculating the distances using that information. But just because you can do it that way, doesn’t mean it’s the most efficient/reliable method.

  28. mostly agree, but seems like even Wolfram Alpha probably uses ML to some extent.
    for example it needs to interpret the query by breaking it into entities and relations, which need to use entity and relationship extraction. These are also not error proof, unless the understanding/interpreting query is done manually by giving exact values to the retrieve information or using math engines.

  29. The connection between GPT and Wolfram Language is already there. If asked ChatGPT can generate proper Wolfram Language – however far from perfect. And, it can understand error messages from Mathematica. I found myself copying the code from ChatGPT to Mathematica and error messages in the opposite direction. Unfortunately, Wolfram Language is not the language ChatGPT understands best, but this can improve

    So, at a moment those mighty AI tools need a stupid human to cut and paste, but they could easily talk to each other directly, already today.

    One would be then able to describe the mathematical problem in words in ChatGPT and get the result immediately.

    Somebody just needs to connect them.

    Wolfram, please… It would really make a difference.

  30. Here is what ChatGPT says to the combination of Wolfram|Alpha & ChatGPT:

    Could wolfram|Alpha and ChatGPT work together for superior performance?

    Yes, Wolfram|Alpha and ChatGPT could potentially work together for superior performance, depending on the specific use case.

    Wolfram|Alpha is an expert in computational knowledge, providing precise and detailed answers to factual questions based on its vast knowledge base and sophisticated algorithms. It excels in fields like mathematics, physics, chemistry, and engineering, where there are clear and well-defined answers based on data and formulas.

    On the other hand, ChatGPT is a language model that is trained to understand and generate human-like text based on patterns and contexts in natural language. It is designed to hold human-like conversations and provide more open-ended responses to a wide range of topics, from casual conversation to complex reasoning and decision-making.

    Together, Wolfram|Alpha and ChatGPT could provide a more powerful and versatile platform for answering questions and solving problems. For example, if a user asks a question that requires a factual answer that can be found in Wolfram|Alpha’s knowledge base, the system could provide a precise answer from Wolfram|Alpha. If the user asks a more open-ended question that requires a nuanced and context-dependent response, the system could generate a more natural and conversational answer from ChatGPT.

    There are already some examples of systems that combine the strengths of both approaches, such as IBM Watson, which uses natural language processing and machine learning to interact with users and provides expert knowledge in various fields, including healthcare and finance.

  31. Outstanding! I found the same thing in my investigation and noted that ChatGPT was consistently giving wrong mathematical answers but I did not go on to find the source of the errors, instead simple chalking it up to the need for an informed reader to proof the work of a subordinate. This insight takes the discussion to an entirely new level of opportunity!