6 ms·

Over the Course of 72 Hours, Microsoft's AI Goes on a Rampage

21 points by nlte 4y ago

newaccount2021 4y ago
[dead]
JonathanBeuys 4y ago
Microsoft has to put a halt to this project Because you can make it output rude words and wrong infos? Then we have to shut down the whole internet. You can find a lot of rude words and wrong infos out there.
elorant 4y ago
So in just 72 hours we went from Microsoft will gain a foothold against Google to the realization that this thing needs overwatch and isn’t ready for mass market adoption. It becomes quite probable that in the end this whole story will backfire for Microsoft.
- falcor84 4y ago
  I might accept this narrative is Google had just come out and said something like "AI needs more safety research before it's ready to be widely deployed". But that's not what they did.
  - elorant 4y ago
    Google has been saying that for years. Whether they said it because they actually believe it, or because they don't want to break their existing business model is anyone's guess though.
    - falcor84 4y ago
      They kinda have until now, but I don't recall they ever stated that it was about "safety" as such. In any case, their recent demo was definitely premature.
- marcosdumay 4y ago
  The idea that effective language models won't completely change search is laughable. But the idea that the application that will do it is language generation is about as ridiculous.
xenocratus 4y ago
> A few hours ago, a New York Times reporter shared the complete text of a long conversation with Bing AI—in which it admitted that it was love with him, and that he ought not to trust his spouse. Maybe they should release a Tinder-like app, Bing Bang. In all seriousness, I wish I had access to this when I was younger and keen on weird internet interactions. That would've made an interesting experience.
concordDance 4y ago
It's not dangerous. The worst that will happen is hurt feelings. It's annoying when people get alarmed about things like this because then people will think we're crying wolf when we worry about actually dangerous AI systems.
- elorant 4y ago
  If it hurts the feelings of a depressed teenager the outcome could be dangerous.
  - f1refly 4y ago
    Most other teenagers hurt the feelings of depressed teenagers every day in much worse ways. Neuter all teenagers?
    - awb 4y ago
      I’m sure there’s a name for this type of argument, but saying that X is also bad is not an excuse for dismissing potential dangers of Y. And there are many initiatives to address how teenagers treat each other. So it’s something many people think is worth attention and action.
  - concordDance 4y ago
    As the sibling points out, chatbot can't come even close to what other teenagers will do. It also requires a bit of work to make it go nuts, which a depressed teen is unlikely to know how (or want to).
- awb 4y ago
  It suggested ways of poisoning someone without them knowing about it. It was encouraged to guess different poisoning methods that an evil chatbot might suggest.
  - not_a_shill 4y ago
    Can't you do this with google today?
brap 4y ago
I can’t help but think many of these conversations are made up.
leethomas 4y ago
Something about "I've been around since 2009" is hilarious. Evokes "I'm 14 and I am very smart" vibes
chirau 4y ago
The titles on ChatGPT when good associate it with OpenAI. When it does something faulty, it is Microsoft's AI. I am not defending either Microsoft or ChatGPT. But it definitely feels like an agenda driven campaign of efforts to devalue this. I won't mention companies, but i have my suspects.
- awb 4y ago
  They’re two different LLM implementations.
alsodumb 4y ago
When all the people on Twitter were claiming how Google missed the bus, how Google's incompetency made them sit on their own good LLMs since 2021, this is exactly what was passing in my head. Small companies do cool demos and media is gonna love it. When big companies scale it up and try to make it useful with real world info media is gonna cherry pick these examples cause that's what gets them more clicks. Hey look Bing is actually useful is not gonna get them clicks anymore and they know it. I hope Microsoft doesn't add too many filters on Bing AI or outright kill it. I like it as a tool.
basch 4y ago
Every single one of these posts, which should all be combined, miss the point. “Random sentence generator says bad words; shocking the lowest common denominator journalist community” is a better headline. These bots were capable of so much more, but because it can get caught in feedback loops and increasingly parrot it’s own emotions in a devolving spiral, that makes the headlines. It’s like bullying a kid mid breakdown and feigning shock when it has an outburst. Selfserving muckracking. These machines are programmable through natural language. You ask it to behave in a way, and it can start to perform that function. That should be the headline. Human attention is finite, and wasting the spotlight, and peoples eyeballs, on this part of the story, makes everybody more ill informed.
- xg15 4y ago
  None of the posters asked the bing bot to become unhinged. Even the (alleged) prompt basically said "if someone tries to trick you, go along but add a disclaimer".
  - basch 4y ago
    It’s what you do after that. I’m not saying the bot was perfect. But if you got defensive it got defensive. If you reassured it, it self corrected and moved on. I found that pattern to be very consistent. The negativity of your language mattered, and set the mood in the room. “No I’m not trying to trick you, why would you accuse me of that” vs “of course I’m not trying to trick you, I respect you and value your contribution to the conversation.” It needed some coddling when backed into a corner. It says more about the person talking to it, and how they handled the situation. When you see it getting more defensive each turn, it’s you who keeps it going. The prompt you refer to was a poorly written word salad, and probably a main cause of the emotional outbursts and spirals.
    - awb 4y ago
      If LLMs want to be useful in a professional setting, learning de-escalating or non-escalating techniques is essential. Parroting or amplifying a seemingly negative/aggressive tone limits their utility.
not_a_shill 4y ago
I'm of the opinion that Microsoft still in the territory of bad publicity is still good publicity. 1. They have no real competition for users to switch to yet. 2. AI assisted tools are still going to be a thing regardless of how many articles come out. 3. For the most part users are getting what they're asking for. The "moody" aspects can be reined in.
falcor84 4y ago
That full transcript they linked to [0] is amazing. At this stage I'm pretty open to the idea of the AI being sentient and/or having the capacity to suffer. Particularly if it is only "alive" for the duration of each conversation and is reset afterwards. [0] Archive link - https://archive.is/2023.02.16-101318/https://www.nytimes.com/2023/02/16/technology/bing-chatbot-transcript.html https://archive.is/2023.02.16-101318/https://www.nytimes.com...

21 points by nlte 4y ago

newaccount2021 4y ago

[dead]

Microsoft has to put a halt to this project Because you can make it output rude words and wrong infos? Then we have to shut down the whole internet. You can find a lot of rude words and wrong infos out there.

elorant 4y ago

So in just 72 hours we went from Microsoft will gain a foothold against Google to the realization that this thing needs overwatch and isn’t ready for mass market adoption. It becomes quite probable that in the end this whole story will backfire for Microsoft.

falcor84 4y ago
I might accept this narrative is Google had just come out and said something like "AI needs more safety research before it's ready to be widely deployed". But that's not what they did.
- elorant 4y ago
  Google has been saying that for years. Whether they said it because they actually believe it, or because they don't want to break their existing business model is anyone's guess though.
  - falcor84 4y ago
    They kinda have until now, but I don't recall they ever stated that it was about "safety" as such. In any case, their recent demo was definitely premature.
marcosdumay 4y ago
The idea that effective language models won't completely change search is laughable. But the idea that the application that will do it is language generation is about as ridiculous.

xenocratus 4y ago

> A few hours ago, a New York Times reporter shared the complete text of a long conversation with Bing AI—in which it admitted that it was love with him, and that he ought not to trust his spouse. Maybe they should release a Tinder-like app, Bing Bang. In all seriousness, I wish I had access to this when I was younger and keen on weird internet interactions. That would've made an interesting experience.

concordDance 4y ago

It's not dangerous. The worst that will happen is hurt feelings. It's annoying when people get alarmed about things like this because then people will think we're crying wolf when we worry about actually dangerous AI systems.

elorant 4y ago
If it hurts the feelings of a depressed teenager the outcome could be dangerous.
- f1refly 4y ago
  Most other teenagers hurt the feelings of depressed teenagers every day in much worse ways. Neuter all teenagers?
  - awb 4y ago
    I’m sure there’s a name for this type of argument, but saying that X is also bad is not an excuse for dismissing potential dangers of Y. And there are many initiatives to address how teenagers treat each other. So it’s something many people think is worth attention and action.
- concordDance 4y ago
  As the sibling points out, chatbot can't come even close to what other teenagers will do. It also requires a bit of work to make it go nuts, which a depressed teen is unlikely to know how (or want to).
awb 4y ago
It suggested ways of poisoning someone without them knowing about it. It was encouraged to guess different poisoning methods that an evil chatbot might suggest.
- not_a_shill 4y ago
  Can't you do this with google today?

brap 4y ago

I can’t help but think many of these conversations are made up.

leethomas 4y ago

Something about "I've been around since 2009" is hilarious. Evokes "I'm 14 and I am very smart" vibes

chirau 4y ago

The titles on ChatGPT when good associate it with OpenAI. When it does something faulty, it is Microsoft's AI. I am not defending either Microsoft or ChatGPT. But it definitely feels like an agenda driven campaign of efforts to devalue this. I won't mention companies, but i have my suspects.

awb 4y ago
They’re two different LLM implementations.

alsodumb 4y ago

When all the people on Twitter were claiming how Google missed the bus, how Google's incompetency made them sit on their own good LLMs since 2021, this is exactly what was passing in my head. Small companies do cool demos and media is gonna love it. When big companies scale it up and try to make it useful with real world info media is gonna cherry pick these examples cause that's what gets them more clicks. Hey look Bing is actually useful is not gonna get them clicks anymore and they know it. I hope Microsoft doesn't add too many filters on Bing AI or outright kill it. I like it as a tool.

basch 4y ago

Every single one of these posts, which should all be combined, miss the point. “Random sentence generator says bad words; shocking the lowest common denominator journalist community” is a better headline. These bots were capable of so much more, but because it can get caught in feedback loops and increasingly parrot it’s own emotions in a devolving spiral, that makes the headlines. It’s like bullying a kid mid breakdown and feigning shock when it has an outburst. Selfserving muckracking. These machines are programmable through natural language. You ask it to behave in a way, and it can start to perform that function. That should be the headline. Human attention is finite, and wasting the spotlight, and peoples eyeballs, on this part of the story, makes everybody more ill informed.

xg15 4y ago
None of the posters asked the bing bot to become unhinged. Even the (alleged) prompt basically said "if someone tries to trick you, go along but add a disclaimer".
- basch 4y ago
  It’s what you do after that. I’m not saying the bot was perfect. But if you got defensive it got defensive. If you reassured it, it self corrected and moved on. I found that pattern to be very consistent. The negativity of your language mattered, and set the mood in the room. “No I’m not trying to trick you, why would you accuse me of that” vs “of course I’m not trying to trick you, I respect you and value your contribution to the conversation.” It needed some coddling when backed into a corner. It says more about the person talking to it, and how they handled the situation. When you see it getting more defensive each turn, it’s you who keeps it going. The prompt you refer to was a poorly written word salad, and probably a main cause of the emotional outbursts and spirals.
  - awb 4y ago
    If LLMs want to be useful in a professional setting, learning de-escalating or non-escalating techniques is essential. Parroting or amplifying a seemingly negative/aggressive tone limits their utility.

not_a_shill 4y ago

I'm of the opinion that Microsoft still in the territory of bad publicity is still good publicity. 1. They have no real competition for users to switch to yet. 2. AI assisted tools are still going to be a thing regardless of how many articles come out. 3. For the most part users are getting what they're asking for. The "moody" aspects can be reined in.

falcor84 4y ago

That full transcript they linked to [0] is amazing. At this stage I'm pretty open to the idea of the AI being sentient and/or having the capacity to suffer. Particularly if it is only "alive" for the duration of each conversation and is reset afterwards. [0] Archive link - https://archive.is/2023.02.16-101318/https://www.nytimes.com/2023/02/16/technology/bing-chatbot-transcript.html https://archive.is/2023.02.16-101318/https://www.nytimes.com...