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

411–420 of 426 posts

Re: Chatbots: Still dumb after all these years

#411

Post by parkingrift »

perryizgr8 wrote: post #97
parkingrift wrote: post #54

Earlier quoted context omitted.

These bots seem equivalent to an adult with amnesia after every spoken sentence. Absolute understanding of the language, and some impressive display of recalling facts, but without any understanding of the environment or context of the conversation. This is polar opposite to any experience I've had with children. Children are aware of their environment and have complex thoughts, but sometimes they are unable to conve…

When you talk to children who haven't developed the skill of talking completely, you still get the sense that there's a lot going on inside that they're unable to express. Sometimes they will show it to you with their actions. I wonder if chat bots are also struggling with the language, but have an internal story that's playing out, desperate to be understood. But they can't, because the only interface they have is a…

I can't speak to that because I honestly have no idea.

...but GPT3 doesn't seem to have some inner monologue. It is just flat unable to recall context or understand the environment.

The only opportunity I had to chat with GPT3 I said hi, and then asked how old it was. GPT3 gave me some answer about how it was created in 2016 but it keeps evolving. I said wow that's pretty young. GPT3 says yes I am young, but learning fast. I then asked GPT3 at what point would it consider itself old. GPT3 says 10 years ago, maybe 20 or 30 years ago. I respond to GPT3 with confusion. GPT3 then starts talking about Bitcoin. Literally.

Re: Chatbots: Still dumb after all these years

#412

Post by anotheryou »

Interestingly there is a big divide between chit-chat bots and those having purpose.

The chit-chat bots at best have "google knowledge", the voice assistants only take small commands and the customer service automation bots are really stiff.

Anyways it's the customer service automation bots than already can do a lot of good. If you manage to automate away the easy portions of chats that's already a big win. Further bots can do many actions really quickly so you have no wait time if it can handle your conversation. I think building a good co-pilot scenario where you hand conversations back and forth to a human at the right moments is the near future here.

Re: Chatbots: Still dumb after all these years

#414

Post by sam0x17 »

systemvoltage wrote: post #324
ClumsyPilot wrote: post #310

Earlier quoted context omitted.

"So the attack on capitalism is so deeply deluded" Under Colonial powers corporations engaged in slave trade, folks that fought against slavery were fighting against capitalism, totally deluded right? Ancient Japan had corporations in year 578 A.D., where they capitalist? Fascism and Nazis had corporations, and they made great scientific progress, were they capitalist? If the answer to any of the above is no, then ma…

I find it hilarious that the strongest response is about some ancient examples of centuries ago. If anything, this is a praise of the system we have today. People don’t understand Capitalism and they seem to have joined the liberal progressive bandwagon. Progressives used to be extremely pro-Capatilist movement. Until lately, and some fringe socialist ideas of Bernie Sanders, it was universally accepted. Apparently a…

You don't seem to understand what liberal means. Liberals are pro status-quo, pro lobbyist, pro deregulation of corporations. They are the centrist monolith that stands in the way of leftists and real progressives at every corner of government. Most liberals, transplanted outside of the U.S., would be right or far right without changing their politics.

Re: Chatbots: Still dumb after all these years

#415

Post by sam0x17 »

systemvoltage wrote: post #324
ClumsyPilot wrote: post #310

Earlier quoted context omitted.

"So the attack on capitalism is so deeply deluded" Under Colonial powers corporations engaged in slave trade, folks that fought against slavery were fighting against capitalism, totally deluded right? Ancient Japan had corporations in year 578 A.D., where they capitalist? Fascism and Nazis had corporations, and they made great scientific progress, were they capitalist? If the answer to any of the above is no, then ma…

I find it hilarious that the strongest response is about some ancient examples of centuries ago. If anything, this is a praise of the system we have today. People don’t understand Capitalism and they seem to have joined the liberal progressive bandwagon. Progressives used to be extremely pro-Capatilist movement. Until lately, and some fringe socialist ideas of Bernie Sanders, it was universally accepted. Apparently a…

The strongest example is the opioid epidemic. Whoops.

Re: Chatbots: Still dumb after all these years

#416

Post by kidproquo »

This is more problematic with the USCIS immigration helpline bot. You don't have the option to yake your business elsewhere.

Trying to get the phone bot to let me talk to a human at USCIS is like talking to the Avrona Kern AI from Children of Time by Adrian Tchaikovsky.

Me: The typographic error submission site throws an error when I try to use it.

Avrona Kern: We have a website to submit typographic errors.

Avrona Kern: Would you like to receive the link via SMS or email?

Me: That site doesn't work! Agent.

Avrona Kern: Would you like to receive the link by SMS or email?

Me: Agent

Avrona Kern: I can help you.

Me: Agent

Avrona Kern: If you keep asking for agent I will end the call.

Me: Typographic site throws error

Avrona Kern: We have a website...

Me: FML

Googling revealed that there are some keywords that make Avrona give up and let me through. Not gonna reveal that here, in case they close those loopholes and some poor schmuck is left with no options.

Re: Chatbots: Still dumb after all these years

#417

Post by sam0x17 »

systemvoltage wrote: post #325
mxkopy wrote: post #301

Earlier quoted context omitted.

You are conflating the organization of human activity with the most dominant form of organization today. Capitalism made all of these things happen today . 50,000 years ago, were equivalent behaviors organized in a similar way? I don't think so. And to think that the space of human behavior is significantly different than it was 50,000 years ago is misguided, in my opinion.

Yeah, for some reason most anti-capitalists cite the bad parts of Capitalism and broad brush the entire concept. If humans can get together and collectively improve the world while feeding their families from the profits, there is no wrong doing. No one hands over profits, it’s earned and it’s a reward from the society through the mechanism of free markets. It’s probably the only thing I’d prefer to do instead of wor…

I'm in no way against capitalism. As I say in my post, the only way forward is capitalism with extreme regulation. Without regulation, capitalism won't get you anywhere good. It should be treated like a bratty child with no moral values and poor upbringing. It has to be dragged, kicking and screaming, towards the good.

Re: Chatbots: Still dumb after all these years

#418

Post by sam0x17 »

simplestats wrote: post #206
sam0x17 wrote: post #143

Earlier quoted context omitted.

Capital is far too short-sighted to look beyond this year or this quarter. Many of the important groundbreaking ideas or projects take years, even decades to formulate. As I've said elsewhere, capital will just take the easiest road it is allowed to take, which currently is stifling innovation, buying competitors, maintaining monopolies, relying on economies of scale to shut out competitors when they can't be acquire…

I cant think of the last time I saw the word "corporations" used in a post that wasn't part of an attack on them as inferior/evil/villanous etc. Speaking of NLP, your post would be very easy to classify sentiment of using a simple keyword approach. Anyway, I've worked both in academia and industry. Industry is simply more practical and better at technology. In tech areas, academia desperately needs industry to provid…

And what changed? We stopped regulating corporations, and they started becoming bad for society. You have to lead corporations to the good, with guardrails, incentives, and regulations, like a petulant child. Without this safety-net in place, you get things like the opioid epidemic.

Re: Chatbots: Still dumb after all these years

#419

Post by p1esk »

Jensson wrote: post #328
fault1 wrote: post #317

Earlier quoted context omitted.

I agree with all that (though I don't really necessarily associate any of that with computer vision, perhaps it's my physics bias). Having good state estimation, kinematic, and dynamics models of real world objects is something that is very mature in controlled environments, but not very mature in other environments.

I think you might over-estimate what the current computer vision models are capable of. They are very good at recognizing what class an object belongs to, but they aren't very good at extracting data about the object based on the image. Extracting data of objects from images is image recognition, and humans and most animals relies heavily on vision to get data about objects.

they aren't very good at extracting data about the object based on the image

I’d say they are pretty good: https://towardsdatascience.com/a-guide-to-image-captioning-e...

Re: Chatbots: Still dumb after all these years

#420

Post by gwern »

katmannthree wrote: post #347
mcswell wrote: post #339

Earlier quoted context omitted.

Apparently they (or at least the person you were responding to was "talking" to) also "know" about data they find on the web, which does not need to be in their training data. There was a news article recently about an 8(?) year old who asked Alexa for a "challenge." "Plug a charger part way into an electric outlet," replied the Alexa", "and touch a penny..." (you can guess the rest--I won't put it here in case anoth…

I don't believe GPT-3 (which is what I was using and commenting on) has access to the internet. It was trained on a large data set from the web, but for instance if you ask it about COVID-19 it has absolutely no idea what you're talking about. Alexa is a different matter entirely, that one actively troll the web in response to your queries.

The default GPT-3 Playground doesn't have web access, no. But that's a mere contingent fact. You can certainly add in retrieval capacities to large language models (that was a very hot trend last year), and even GPT-3 can browse the web - you may have seen a demo or 2 on Twitter, but OA did it much more seriously recently with https://openai.com/blog/improving-factual-accuracy/ Letting GPT-3 browse a corpus of web pages, picking what links to follow, summarizing, and using them to generate an answer.

(Just one of the many capabilities GPT-3 has been shown to be capable of, which OP will never tell you about, because he's too busy using it in the dumbest way possible and peddling misinformation about eg. what sampling is for - no, Dr Professor Smith, temperature sampling is not about 'avoiding repetition', which is good, because it doesn't avoid repetition anyway...)

Post reply on HN