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

Chatbots: Still dumb after all these years

mindmatters.ai

Post reply on HN
Print view View original

221–230 of 426 posts

Re: Chatbots: Still dumb after all these years

#221

Post by Sebb767 »

hdjjhhvvhga wrote: post #73
pwinnski wrote: post #13

Having worked in ML at two different companies now, I think that people interpreting model output as intelligence or understanding says much more about the people than about the model output. We want it to be true, so we squint and connect dots and it's true. But it isn't. It's math and tricks, and if human intelligence is truly nothing more than math and tricks, then what we have today is a tiny, tiny, tiny fraction…

> Having worked in ML at two different companies now, I think that people interpreting model output as intelligence or understanding says much more about the people than about the model output. I'd add it says a lot about all the companies that advertise them in a way that has nothing to do with reality, and those who actually buy them. We all know the usefulness of these things is below zero, because they only get i…

Currently, the way Chatbots are, they are just some glorified text boxes to enter your information in, a different format of the search box and lastly, a way to make sure you've done the basic checklist. They also hardly speak your language and refuse to deviate from the script in any way. Which, without much irony, is pretty much my experience with the first level support at a lot of companies. So I'd say they were quite successful with replacing that.

Re: Chatbots: Still dumb after all these years

#222

Post by laurensr »

What struck me is that the good solutions are trained on non-English languages. When I contacted these companies about offering their service in English or my native language, the best they could come up with is putting something like Google Translate in front of their solution.

Re: Chatbots: Still dumb after all these years

#223

Post by AussieWog93 »

jazzyjackson wrote: post #204
jhbadger wrote: post #117

Earlier quoted context omitted.

>who is the current president? I wonder what happened to IBM Watson's technology that was good at answering trivia -- it actually won a game of Jeopardy! a decade ago (although of course it answered things with a question as per the rules). I know that they weren't that successful at applying it to biomedical research as they had hoped, but it would seem it would be better at chatbots than GPT-3 and other deep-learni…

I suppose it didn't work with less than 15 terabytes of RAM

For anyone wondering, this is not an exaggeration. Watson actually had 15 (or 16, depending on the source) TB of RAM, as well as a 2880-core processor.

Re: Chatbots: Still dumb after all these years

#224

Post by CRConrad »

300bps wrote: post #89
mithr wrote: post #84

Earlier quoted context omitted.

I think things started to go wrong here > The knight then grabs his longsword and bows. "bow" here probably refers to "a bending of the head or body in respect, submission, assent, or salutation", rather than to the weapon. The knight grabbed his longsword and bowed to you. Your confusion then confused the game. This does expose an interesting common "failure mode" for these types of bots: they assume that all conver…

Not a bad theory but the very first thing I did was enter "look" and it did say there was a longsword on the ground and a bow on the table. Here is the entire unedited transcript: "Good luck, my friend!" the man says. You inventory. You have a short sword, a longsword, a shield, and a crossbow with bolts. You look. You look around and see a longsword lying on the ground and a bow on a table. A large dragon looms in t…

Yes -- "the knight then grabS his longsword and bowS.". That 's' at the end there indicates present tense of the verb "to grab" and "to bow", just like in "roars and swipes". (I guess for knights it's just good manners to thank by bowing when they're gifted a sword.) It's not a plural marker of a noun, because there was only one bow on the ground.

Re: Chatbots: Still dumb after all these years

#225

Post by Jackson__ »

canjobear wrote: post #140

The author didn't prompt GPT-3 properly. I gave it the prompt below: > The following is a conversation between a human and an AI assistant. The AI assistant is helpful, friendly, and intelligent. > Human: Good morning! I'm going to ask you a few questions. > AI: Alright, I'll do my best to answer them. > Human: Which is bigger, Mt. Everest or a shoebox? > AI: Mt. Everest. > Human: [question] > AI: And I got these ans…

I agree that the author does not appear to have done much of any basic research regarding this topic. It reminds me of articles that were published just after GPT3/GPT2 were released, while also not bringing any unique perspective to the topic.

It is saddening me considering it is 2022 now, nearing 2 years after the release of GPT3. Yet journalists retreat the same ground, pretending to have something insightful to say, after what appears to be a 30 minute session with OpenAI's GPT3 Sandbox.

Re: Chatbots: Still dumb after all these years

#227

Post by TedShiller »

pwinnski wrote: post #13

Having worked in ML at two different companies now, I think that people interpreting model output as intelligence or understanding says much more about the people than about the model output. We want it to be true, so we squint and connect dots and it's true. But it isn't. It's math and tricks, and if human intelligence is truly nothing more than math and tricks, then what we have today is a tiny, tiny, tiny fraction…

The reason laypeople want it to be true is because experts present it as being true.

Re: Chatbots: Still dumb after all these years

#228

Post by aidenn0 »

dane-pgp wrote: post #189
hotpotamus wrote: post #23

Earlier quoted context omitted.

I'm not trying to be mean or break any HN rules, but did you read the article? It basically covers what you asked and was quite a revelation to me. Others in here point out that GPT-3 is not a chatbot which is good info, but I also wonder if there is anything out there that can even "remember" the context of a conversation it's supposed to be having. Particularly interesting is the question from the article, "who is…

> this is a common question paramedics will give to people who have suffered head injuries to assess consciousness. I can't help wondering what the assessment would be if someone answered "Who is the current president?" with "Watson Bakery". https://www.youtube.com/embed/cS1Tmz816hk

To be fair, if I were asked the current PM of Japan the only name I could come up with is Abe, and that's only because he was PM for so very long.

Re: Chatbots: Still dumb after all these years

#229

Post by dole »

lijogdfljk wrote: post #159
banana_giraffe wrote: post #127

In the early days I worked at a company that had a natural language chatbot product. It wasn't an online thing, but rather part of a larger tool. You could ask it to do things like "show me the quarterly spreadsheet", and if it didn't know what "quarterly spreadsheet" was, it'd ask questions in English and learn what "quarterly" and "spreadsheet" meant. And it could use that new knowledge to update its questions so i…

Man that sounds like a ton of fun to make for an at-home chat bot. I used to love designing these sorts of apps for no real purpose lol.

Mine suggests a personalized meal from Taco Bell.

"You should get a Potato Waffle Taco Party Pack, a Grilled Crunchwrap Slider Quesadilla and a Smothered Cool Ranch® Doritos® Locos Power Bowl."

Re: Chatbots: Still dumb after all these years

#230

Post by aidenn0 »

dexwiz wrote: post #45

Chatbots are about as useful as phone trees. They can help solve the top 5 easy/common problems, but they are useless passed that. Anyone who has worked in a call center knows that more than half the calls are about the same couple of issues: reset password, finding common help docs, etc. Since help desks are cost centers, it makes sense to half a robot handle as many of these as possible. I think most of the hate di…

Just please give me a DTMF tree as a fallback for voice recognition. I was trying to make a change to a rental-car reservation from a noisy airport, and I spent 15 minutes trying to get through the voice-recognition chatbot...
Post reply on HN