ChatGPT can now call Wolfram Alpha • Hacker News Board

Re: ChatGPT can now call Wolfram Alpha

#311

Post by TrevorJ » Fri, Mar 24, 2023, 6:18 PM UTC

xdennis wrote: post #240

sharkweek wrote: post #182

Earlier quoted context omitted.

Please anyone willing to answer as if you might be speaking to a golden retriever: How important is a "lead" in this space? I have been comparing ChatGPT and Bard a lot the last two days because this is all very fascinating, and from my datapoint-of-one perspective, it feels like ChatGPT is way, way ahead. Then I start to wonder is this because ChatGPT captivated our attention so much the last few months that they've…

I don't think OpenAI's lead is holding Google back. Companies are held back by their own mentalities, and newer companies have fewer restrictions. As an example, Google supposedly had better text-to-image models than DALL-E/Midjourney/etc, but didn't release them because they "reinforced harmful stereotypes". It's also probably holding back on LLM because it doesn't want to harm its search cash cow, just like Kodak d…

I think large companies have a huge disadvantage in this area due to reputational risk that the startups don't have to really worry about to the same extent.

Re: ChatGPT can now call Wolfram Alpha

#312

Post by TrevorJ » Fri, Mar 24, 2023, 6:37 PM UTC

antibasilisk wrote: post #64

kingstoned wrote: post #45

Earlier quoted context omitted.

Judging by how underwhelming Bard is, I think they are toast, but at least they have YouTube: https://twitter.com/paulg/status/1620539538476249088

YouTube is a glut of audio-visual training data they have the copyright to, I wouldn't be surprised if they started training something to replace fixed length videos entirely. The content creator will become obsolete, which is tragic.

Eh, one of the primary value props of good youtube channels is the authenticity. I think this will be a tougher nut to crack for AI than most.

Re: ChatGPT can now call Wolfram Alpha

#313

Post by bacchusracine » Fri, Mar 24, 2023, 7:33 PM UTC

So ChatGPT is now at the 'Who wants to be a Millionaire' stage and can phone a friend?

Re: ChatGPT can now call Wolfram Alpha

#314

Post by cubefox » Fri, Mar 24, 2023, 9:17 PM UTC

I'm more skeptical than most people here.

Except for certain clearly defined math problems (even math problems often aren't like that), WolframAlpha has turned out to be surprisingly useless for me in the last years. All these examples Wolfram mentions in his blog were possible via the WolframAlpha website/app before. Granted, now you don't have to worry about getting your WA query exactly right (a major problem in the past), since the GPT will handle this for you. But most of these examples seem still of little value to the vast majority of people.

Distance between Chicago and Tokyo? Planetary Moons larger than Mercury? Those examples are typical for WolframAlpha in that they are of very little interest.

I predict that in a few weeks most people here will have mostly forgotten about the Wolfram plugin, as it will turn out to be rarely useful. The plugins by OpenAI themselves (web access, compiler) are much more promising.

Re: ChatGPT can now call Wolfram Alpha

#315

Post by SideQuark » Sat, Mar 25, 2023, 2:16 AM UTC

bottlepalm wrote: post #174

ElijahLynn wrote: post #120

Last paragraph: I see what’s happening now as a historic moment. For well over half a century the statistical and symbolic approaches to what we might call “AI” evolved largely separately. But now, in ChatGPT + Wolfram they’re being brought together. And while we’re still just at the beginning with this, I think we can reasonably expect tremendous power in the combination—and in a sense a new paradigm for “AI-like co…

Between this and his last post, he continues to dismiss ChatGPT as fancy next word statistics.

That's exactly what ChatGPT is. It's specifically trained as a next word (well, token) predictor. It has no long term plan, only a local window of context. And inference is completely predict the next token, with a large scales stats prediction model.

Re: ChatGPT can now call Wolfram Alpha

#316

Post by Mike_12345 » Sat, Mar 25, 2023, 4:44 AM UTC

RomanPushkin wrote: post #12

Looks like "How to use MS-DOS superpowers on your MacOS".

Then you don't understand what Wolfram Alpha is capable of. It is far superior to language models for mathematics. No comparison. That's why they are integrating it.

Re: ChatGPT can now call Wolfram Alpha

#317

Post by Melingo » Sat, Mar 25, 2023, 8:16 AM UTC

ElevenLathe wrote: post #302

Melingo wrote: post #281

Earlier quoted context omitted.

And that was the same thing as maps? Free? Global

Yes, it was free and global. The main things that Google Maps had over MapQuest were a nice AJAXy interface and (eventually) street view. But MapQuest was itself basically a generational improvement on AAA's "TripTiks."

Never used it.

And it was doing also routing too? Also in Germany?

Re: ChatGPT can now call Wolfram Alpha

#318

Post by childintime » Sat, Mar 25, 2023, 12:47 PM UTC

jacquesm wrote: post #181

TOMDM wrote: post #6

Well, this shores up one of ChatGPT's biggest weaknesses, especially GPT4's. I wonder when we'll see the pace of refinement for this drop off.

If wonder if it will drop off.

Right, it will already be out of our view when it does.

Re: ChatGPT can now call Wolfram Alpha

#319

Post by bottlepalm » Sat, Mar 25, 2023, 6:22 PM UTC

SideQuark wrote: post #315

bottlepalm wrote: post #174

Earlier quoted context omitted.

Between this and his last post, he continues to dismiss ChatGPT as fancy next word statistics.

That's exactly what ChatGPT is. It's specifically trained as a next word (well, token) predictor. It has no long term plan, only a local window of context. And inference is completely predict the next token, with a large scales stats prediction model.

The long term plan is what generates the next word.

Re: ChatGPT can now call Wolfram Alpha

#320

Post by tpmx » Sat, Mar 25, 2023, 9:30 PM UTC

Enginerrrd wrote: post #309

hathawsh wrote: post #52

Earlier quoted context omitted.

I also liked Transcendence, but you have to admit the path to AGI was vastly oversimplified. No one is creating AI by uploading their brain into a computer.

I actually really disagree. I think that's a VERY plausible way for a true AGI to come about. In fact, prior to LLM's I used to think that was the ONLY way we were going to get an AGI. We know humans are AGI's. If it were possible to upload your brain to a computer, and then make modifications to enhance your speed of thought and intelligence, of course you would do that. It actually makes a LOT of sense if you'll ad…

Yeah. A better counterpoint would have been "but your brain scanning tech is subpar", to which I say phooey, it's a fiction movie.