This couldn’t be a more definitive comparison of what ChatGPT isn’t good at. WolframAlpha was released in 2009 but why isn’t it getting the same sexy accolades and valuation as ChatGPT?
Wolfram Alpha and ChatGPT
71–80 of 309 posts
Re: Wolfram Alpha and ChatGPT
#72Earlier quoted context omitted.
"Also the Contact Us thing is a red flag for me." Why is that a red flag and not something you see as potentially finding a custom solution to your need?
Not OP, but I feel similarly and for me it's because the information assymmetry is a deliberate way to give advantage to the seller. I also perceive deliberate lack of transparency as dishonest and exploitative. If you can't give me a price upfront, I probably don't want to do business with you. That may not be entirely fair, but that is my perception I also don't like doing business with people I can't trust on a ha…
"If you can't give me a price upfront, I probably don't want to do business with you."
I don't understand. They give very clear prices for various use cases and for everything else, there is the "contact us" button. Which is quite common as far as I know.
Re: Wolfram Alpha and ChatGPT
#73Earlier quoted context omitted.
google can't calculate to different unit types like wolfram alpha. For example "1.5 cup of uranium in kg" is something wolfram alpha can calculate but nothing else can without extra manual steps.
I don't even want to know what kind of cookies you're baking
They're simply radiant
Re: Wolfram Alpha and ChatGPT
#74In the past, I have found Stephen Wolfram's air of superiority off-putting[a], but in this case I find myself nodding in agreement with every point he makes in the OP. I highly recommend you read it. This proposal, in particular, sounds like a great idea for improving ChatGPT in the near term: > ...there’s the immediate opportunity of giving ChatGPT computational knowledge superpowers through Wolfram|Alpha. So it can…
I agree that the prospects of combining the two is very appealing. I do hope that will happen in one way or another. As for Stephen Wolfram, maybe it's my predisposition but even in this article, I feel like I am sensing not just vanity but also a slight jealousy of ChatGPT's success.
I've just come to accept Stephen Wolfram's "air" of superiority to just that. If you parse it away as a vibe, he is a genuinely curious and invested person, that has stuck to their peculiar vision for more than 3 decades, and puts his money where his mouth is. Every design session of the wolfram language is streamed on youtube, which is pretty incredible to watch. I'm sad that the pricing / licensing makes wolfram products so hard if not impossible to shill in a professional setting, because I think it is a genuinely unique product in the sea of overall uninspired and mediocre takes on what computing could be.
The wolfram conference is one of the most fun conferences I attend.
Re: Wolfram Alpha and ChatGPT
#75Is there any prior work on how to integrate a LLM with something as primitive as a simple calculator? I suspect it's hard. Does ChatGPT even know what question to ask? Integrating LLM's with logic/reasoning/computation seems important, but I wonder if it's a ways off? I doubt anyone would bother integration with Wolfram except for Wolfram, though.
As far as I recall Meta's Galactica has special tokens to mark working out which can include python code and can run the code during inference to get a result.
Re: Wolfram Alpha and ChatGPT
#76This couldn’t be a more definitive comparison of what ChatGPT isn’t good at. WolframAlpha was released in 2009 but why isn’t it getting the same sexy accolades and valuation as ChatGPT?
Because it only understands language to the extent needed to do math, and that's the easy part.
Merge WolframAlpha and ChatGPT and give it real-time access to the Web, and then things will get interesting.
Re: Wolfram Alpha and ChatGPT
#77Is there any prior work on how to integrate a LLM with something as primitive as a simple calculator? I suspect it's hard. Does ChatGPT even know what question to ask? Integrating LLM's with logic/reasoning/computation seems important, but I wonder if it's a ways off? I doubt anyone would bother integration with Wolfram except for Wolfram, though.
What you're looking for is an action transformer. https://www.adept.ai/act
Re: Wolfram Alpha and ChatGPT
#78> 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…
Great comment! I didn't see Wolfram's OP as being bearish on the "naive" scaling hypothesis, but more as him being a bit upset and jealous about ChatGPT recently getting so much attention. Maybe he's a bit scared too about what future improvements to ChatGPT-like models could do to his business. In any case, I think his proposal of combining the two approaches makes a lot of sense in the near term , as I wrote here:…
> him being a bit upset and jealous about ChatGPT recently getting so much attention.
Was it given away by the fact that he couldn't wait a single paragraph before claiming to have not one but two successes on the scale of ChatGPT? Stephen Wolfram is consistent if anything.
Re: Wolfram Alpha and ChatGPT
#79I'm surprised this isn't on the front page.
It is on the front page now. The algorithms of HN appear to downrank some topics that appear too frequently until they get sufficient positive engagement to avoid filling up the front page with just one thing (which gets boring).
Have they ever publicly discussed what their algorithm is? I'd be quite interested to hear from a place with reasonably high traffic how they go about it.
Re: Wolfram Alpha and ChatGPT
#80This just in, Microsoft in talks to buy Wolfram Alpha.