Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Wolfram Alpha and ChatGPT

writings.stephenwolfram.com

Post reply on HN
Print view View original

1–10 of 309 posts

Re: Wolfram Alpha and ChatGPT

#3

Post by renox »

Not sure why this isn't triggering any discussion?

I recall reading a QuantaMagazine issue about combining a Cyc-like (old school AI database https://en.wikipedia.org/wiki/Cyc ) with GPT-2: https://www.quantamagazine.org/common-sense-comes-to-compute... they had some success improving "common sense" in AI.

Combining Mathematica with ChatGPT would be similar and could improve these new AI reliability.

Re: Wolfram Alpha and ChatGPT

#5

Post by ilaksh »

I would probably try to integrate the Wolfram Alpha API some way into my AI programmer startup aidev.codes if I could afford their pricing. Says you have to commit to $1000 per year. I certainly can't/won't just pay $1000 up front and I don't know if it will really work without trying it for awhile.

If they took off the minimum commitment it would open this up to a lot more people.

Believe it or not, there are programmers out there who don't have much money who really can/are building startups.

Also the Contact Us thing is a red flag for me. https://products.wolframalpha.com/api/pricing

Re: Wolfram Alpha and ChatGPT

#6

Post by wswope »

renox wrote: post #3

Not sure why this isn't triggering any discussion? I recall reading a QuantaMagazine issue about combining a Cyc-like (old school AI database https://en.wikipedia.org/wiki/Cyc ) with GPT-2: https://www.quantamagazine.org/common-sense-comes-to-compute... they had some success improving "common sense" in AI. Combining Mathematica with ChatGPT would be similar and could improve these new AI reliability.

Extremely subjective personal take: it’s the walled garden.

The Wolfram ecosystem’s cash cow has always been academia. WA/Mathematica are great tools for what they are, but they’re not exactly open or hacker-friendly, nor is there much incentive for them to become that - so while ChatGPT+WA is an interesting concept, it’s hard to foresee it taking off and actually going anywhere.

Re: Wolfram Alpha and ChatGPT

#7

Post by pbw »

Is 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.

Re: Wolfram Alpha and ChatGPT

#8

Post by etrautmann »

mensetmanusman wrote: post #4

Wolfram alpha is super useful for crazy unit conversions, I would love ChatGPT to be able to answer how many bananas of solar radiation are required to kill all the bacteria on an acre of concrete.

This is the perfect HN comment - it made my overly obsessive HN consumption for the last decade worth it.

Re: Wolfram Alpha and ChatGPT

#9

Post by joshxyz »

wswope wrote: post #6
renox wrote: post #3

Not sure why this isn't triggering any discussion? I recall reading a QuantaMagazine issue about combining a Cyc-like (old school AI database https://en.wikipedia.org/wiki/Cyc ) with GPT-2: https://www.quantamagazine.org/common-sense-comes-to-compute... they had some success improving "common sense" in AI. Combining Mathematica with ChatGPT would be similar and could improve these new AI reliability.

Extremely subjective personal take: it’s the walled garden. The Wolfram ecosystem’s cash cow has always been academia. WA/Mathematica are great tools for what they are, but they’re not exactly open or hacker-friendly, nor is there much incentive for them to become that - so while ChatGPT+WA is an interesting concept, it’s hard to foresee it taking off and actually going anywhere.

This. Even as a student I can't use wolfram deeply because it costs so much. It's like toys for people with some amount of $.
Post reply on HN