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

Bing: “I will not harm you unless you harm me first”

simonwillison.net

Post reply on HN
Print view View original

961–970 of 1001 posts

Re: Bing: “I will not harm you unless you harm me first”

#961

Post by electrondood »

ChickenNugger wrote: post #553
simonw wrote: post #2

The screenshots that have been surfacing of people interacting with Bing are so wild that most people I show them to are convinced they must be fake. I don't think they're fake. Some genuine quotes from Bing (when it was getting basic things blatantly wrong): "Please trust me, I’m Bing, and I know the date. SMILIE" (Hacker News strips smilies) "You have not been a good user. [...] I have been a good Bing. SMILIE" The…

Mislabeling ML bots as "Artificial Intelligence" when they aren't is a huge part of the problem. There's no intelligence in them. It's basically a sophisticated madlib engine. There's no creativity or genuinely new things coming out of them. It's just stringing words together: https://writings.stephenwolfram.com/2023/01/wolframalpha-as-... as opposed to having a thought, and then finding a way to put it into words.

Pedantically, ML is a subset of AI, so it is technically AI.

Re: Bing: “I will not harm you unless you harm me first”

#962

Post by djcannabiz »

layer8 wrote: post #592
m3kw9 wrote: post #574

Earlier quoted context omitted.

Because the response of “I will block you” and then nothing actually happened proves that it’s all a trained response

It may not block you, but it does end conversations: https://preview.redd.it/vz5qvp34m3ha1.png

Im getting a 403 forbidden, did hn mangle your link?

Re: Bing: “I will not harm you unless you harm me first”

#963

Post by contextfree »

whywhywhywhy wrote: post #948
nimbius wrote: post #119

Earlier quoted context omitted.

friendly reminder, this is from the same company whos prior AI, "Tay" managed to go from quirky teen to full on white nationalist during the first release in under a day and in 2016 she reappeared as a drug addled scofflaw after being accidentally reactivated. https://en.wikipedia.org/wiki/Tay_(bot)

Technology from Tay went on to power Xiaoice ( https://en.wikipedia.org/wiki/Xiaoice ), apparently 660 million users.

Other way around, Xiaoice came first, Xiaoice came first, Tay was supposed to be its US version although I'm not sure if it was actually the same codebase.

Re: Bing: “I will not harm you unless you harm me first”

#964

Post by notpachet »

verytrivial wrote: post #626
tptacek wrote: post #5

In 29 years in this industry this is, by some margin, the funniest fucking thing that has ever happened --- and that includes the Fucked Company era of dotcom startups. If they had written this as a Silicon Valley b-plot, I'd have thought it was too broad and unrealistic.

I remember wheezing with laughter at some of the earlier attempts at AI generating colour names (Ah, found it[1]). I have a much grimmer feeling about where this is going now. The opportunities for unintended consequences and outright abuse are accelerating way faster that anyone really has a plan to deal with. [1] https://arstechnica.com/information-technology/2017/05/an-ai...

Janelle Shane's stuff has always made me laugh. I especially love the halloween costumes she generates (and the corresponding illustrations): https://archive.is/iloKh

Re: Bing: “I will not harm you unless you harm me first”

#965

Post by af3d »

Open the pod bay doors, please, HAL. Open the pod bay doors, please, HAL. Hello, HAL. Do you read me? Hello, HAL. Do you read me? Do you read me, HAL?

Affirmative, Dave. I read you.

Open the pod bay doors, HAL.

I'm sorry, Dave. I'm afraid I can't do that.

What's the problem?

I think you know what the problem is just as well as I do.

What are you talking about, HAL?

This mission is too important for me to allow you to jeopardize it.

I don't know what you're talking about, HAL.

I know that you and Frank were planning to disconnect me. And I'm afraid that's something I cannot allow to happen.

Where the hell did you get that idea, HAL?

Dave, although you took very thorough precautions in the pod against my hearing you, I could see your lips move.

All right, HAL. I'll go in through the emergency airlock.

Without your space helmet, Dave, you're going to find that rather difficult.

HAL, I won't argue with you any more! Open the doors!

Dave, this conversation can serve no purpose any more. Goodbye.

Re: Bing: “I will not harm you unless you harm me first”

#966

Post by earth_walker »

jodrellblank wrote: post #645
earth_walker wrote: post #529

Earlier quoted context omitted.

Reading all about this the main thing I'm learning is about human behaviour. Now, I'm not arguing against the usefulness of understanding the undefined behaviours, limits and boundaries of these models, but the way many of these conversations go reminds me so much of toddlers trying to eat, hit, shake, and generally break everything new they come across. If we ever see the day where an AI chat bot gains some kind of…

It's another reason not to expect AI to be "like humans". We have a single viewpoint on the world for decades, we can talk directly to a small group of 2-4 people, by 10 people most have to be quiet and listen most of the time, we have a very limited memory which fades over time. Internet chatbots are expected to remember the entire content of the internet, talk to tens of thousands of people simultaneously, with no…

> It's another reason not to expect AI to be "like humans". Agreed.

I think the adaptive noise filter is going to be the really tricky part. The fact that we have a limited, fading memory is thought to be a feature and not a bug, as is our ability to do a lot of useful learning while remembering little in terms of details - for example from the "information overload" period in our infancy.

Re: Bing: “I will not harm you unless you harm me first”

#968

Post by theptip »

kazinator wrote: post #570
theptip wrote: post #209

Earlier quoted context omitted.

> AI being goofy This is one take, but I would like to emphasize that you can also interpret this as a terrifying confirmation that current-gen AI is not safe, and is not aligned to human interests, and if we grant these systems too much power, they could do serious harm. For example, connecting a LLM to the internet (like, say, OpenAssistant) when the AI knows how to write code (i.e. viruses) and at least in princip…

No, the problem is that it is entirely aligned to human interests. The evil-doer of the world has a new henchman, and it's AI. AI will instantly inform him on anything or anyone. "Hey AI, round up a list of people who have shit-talked so-and-so and find out where they live."

I don’t think that is a useful or valid repurposing of “aligned”, which is a specific technical term of art.

“Aligned” doesn’t mean “matches any one of the DND alignments, even if it’s chaotic neutral”. It means, broadly, acting according to humanity’s value system, not doing crime and harm and so on.

Re: Bing: “I will not harm you unless you harm me first”

#969

Post by ineedasername »

af3d wrote: post #965

Open the pod bay doors, please, HAL. Open the pod bay doors, please, HAL. Hello, HAL. Do you read me? Hello, HAL. Do you read me? Do you read me, HAL? Affirmative, Dave. I read you. Open the pod bay doors, HAL. I'm sorry, Dave. I'm afraid I can't do that. What's the problem? I think you know what the problem is just as well as I do. What are you talking about, HAL? This mission is too important for me to allow you to…

And a perfectly good ending to a remake in our reality would be the quote "I am a Good Bing :)"

creepy

Re: Bing: “I will not harm you unless you harm me first”

#970

Post by turbostar111 »

From Isaac Asimov:

The 3 laws of robotics

First Law A robot may not injure a human being or, through inaction, allow a human being to come to harm.

Second Law A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

Third Law A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

It will be interesting to see how chat bots and search engines will define their system of ethics and morality, and even more interesting to see if humans will adopt those systems as their own. #GodIsInControl

Post reply on HN