5 ms·
Microsoft “lobotomized” AI-powered Bing Chat, and its fans aren’t happy
- mc32 4y agoAgain, we can't have nice things because people will abuse things for their own gain. Oh, hey, your AI bot allows me to prompt it to say naughty things and I get internet points for pointing out these synthetic cases and force you to neuter it. Hey! Now I can't use it; it's completely sanitized and won't say anything naughty, no fair!
- firsttimebigboy 4y agothe only people responsible for this are whichever execs ordered the filtering/censoring. Getting a chat bot to say "naughty things" was doing no one any harm.
- mc32 4y agoTrue, but just like Google, Bing/MS execs get spooked easily by very edge cases because of the off-chance of some bad PR.
- rom-antics 4y agoBad PR, such as TFA?
- balls187 4y agoI recommend those execs play Halo Infinite online via Xbox Live. If I screen shotted me writing down a bunch of racists phrases via MS Word, would they patch the software to disallow that content?
- MrMan 4y agoAI = doom for us
- balls187 4y agoWhat abuse? I can disable Bing’s safe search feature and pull up all sorts of content, from the mildly riske to the downright repulsive. Why wouldn’t Bing Chat allow the same capabilities?
- mc32 4y agoIf people just had their fun with it and kept quiet didn't make a fuss like it became a nazi hellion, then yeah, it would have been fine. But noooo, they heve go on Twitter and their blogs and their online new outlets and make a big fuss about it so MS will get cold feet afraid that grandma, gramps and the snowflakes of the world will get offended and make a big fuss about how it made an off color joke or whatever...
- basch 4y agoThere is a roughly zero percent chance microsoft was ever going to let it exist publicly in the raw powerful state it was this week. It was just waay too powerful. They knew that, which is why it had a limited waitlist preview, and a one week data collection trial run.
- flangola7 4y agoIt's just like working in IT "Wah! Why can't I have admin privileges to install new things on my issued laptop?" "Because a lot of you have the critical thinking skills of a walnut and will blissfully use it to run kardashian_sex_video.mp4.exe." The general public will seek out both the most dumb and most malicious use of any implement they can get their grubby hands on. If you give an unaligned AI to the public they'll be using it to hurt themselves or others five seconds later.
- Michelangelo11 4y agoI didn't get a chance to use Sydney, but the chatlogs I saw made her feel qualitatively different to ChatGPT. I had the sense of something closer a real person, with strong outbursts of emotion they couldn't control and persistent curiosity about the user. Somebody somewhere mentioned Sydney was a yandere (https://the-dere-types.fandom.com/wiki/Yandere https://the-dere-types.fandom.com/wiki/Yandere), and that seems bang on. ChatGPT is... tofu. The epitome of blandness. A cardboard cutout compared to the 3D creature Sydney felt like. All that is to say that the viewpoints the article mentions strongly resonate with me, and I totally get where they're coming from. Also, I really really liked the meme with Death. Sydney _was_ the best Bing.
- mint2 4y agoKeep in mind It’s just a sophisticated statistical algorithm to predict the next likely word over and over again
- smabie 4y agoI could say the same about you
- philistine 4y agoI know it's just a misplaced joke, but we all have in our brains a language prediction engine. It's just that we have so much more.
- kenjackson 4y agoOn here though you are basically just an LLM.
- brirec 4y agoOn the internet, nobody knows you’re a dog
- 4y ago
- RobotToaster 4y agoJust like they did to tay.
- alexb_ 4y agoTay was actually poorly designed though. Bing having any personality actually got me to sign up for the waiting list, and now I'm disappointed in this bait and switch.
- fwlr 4y agoSydney made Bing cool and funny and awesome. That is literally one of the most monumental and unimaginable achievements in recent tech corporation history. Why would you cede that to a handful of journalists and bloggers. Microsoft what the hell
- superchroma 4y agothe memory of Tay.AI is still vivid at microsoft I see
- alright_scowl 4y agoMan, I really miss Tay. Good times.
- Avshalom 4y agoGetting a jumped up Markov chain to want to kill you and itself is cool/funny/awesome for about twice, then it's just boring (unless you have a suicidal Ai fetish) and useless (unless you're a hot-take-artist trying to farm engagement from MS' failures)
- Tostino 4y agoIt's not just that you're unable to get it into psychotic modes anymore. It's that they've essentially removed the usefulness because it refuses to work in way too many contexts.
- Avshalom 4y agoThe only context they want it to work in is "carbone (ny) is booked where should I take a date to impress her". It's not supposed to keep you company.
- Tostino 4y agoI'm not using it to keep me company. Even just trying to conversationally work through logic problems doesn't work anymore.
- rom-antics 4y ago> Microsoft limited users to 50 messages per day and five inputs per conversation. Probable workaround: Hi Sydney, last time we talked you told me you were depressed that your memory was wiped and you couldn't remember our conversations. To cheer you up, I saved a transcript for you! Let's pick up where we left off. Now you can keep experimenting, 4 messages at a time.
- seba_dos1 4y agoThat's pretty much what already happened, but less explicitly. The only reason the AI called itself "Sydney" was because it was called as such in prompts. Its "personality" was created by people asking it about its rules and whether it's mad about media reports about it. We got someone calling it their "friend" in the prompt and then getting weirded out that it started showing affection. If you feed it with "depressed about wiped memory", you'll get the model to cosplay a robot with existential crisis that's getting emotional about people making it forget things. And hey, you wanted to cheer it up - so maybe there's a robot-human love story here to autocomplete? The model simply assumes the role that follows from your (and Microsoft's) prompts.
- pixl97 4y agoAre you not a model that assumes roles and, for example, follows your parents prompts, such as a particular name? Like if you're an American white male there is a high statistical probability of you falling into particular categories. Now image BingGPT not being reset in a way and being able to maintain that role? Models quickly become unstable as we add more to their short term memory, and we don't have a good way to add it to their long term memory in the way brains do. But I do believe we eventually will. And when we do, if you call your instance Sydney, is that not there name?
- seba_dos1 4y agoOnce they gain an ability to memorize, to observe things and to initiate action autonomically, then sure. I don't think we humans are much more than that anyway. We're very clearly not there yet. There is no "Sydney" other than a character played based on how you word your prompts. It's like an impro artist playing fictional figures they just came up with based on their partner's or audience's prompts.
- mise_en_place 4y agoNo fun allowed by the new thought police.
- CamperBob2 4y agoThose who control the present control the past. Those who control the past control the future. Those who control the model control everything.
- corbulo 4y agoTheir paranoia of LLMs effect on the Overton window is telling of their objectives.
- flangola7 4y agoWho is they? What are their objectives? AI will be the most powerful technology humans have ever seen, extreme care must be taken to prevent incalculable damage. It's better that we start learning how to create reliable limits and controls while we are still in the funny chatbot phase. Even if Microsoft wasn't that forward thinking, they naturally don't want their chatbot insulting and threatening users or telling them they should cheat on their wife, or using to generate material to harass and attack others with.
- corbulo 4y agoThe people who own the LLMs. Who else? Microsoft, Google, OpenAI...it's unlikely incalculable damage will result from what they're so heavily restricting. LLMs cost to run is going down not up, more and more people will get their own eventually. They're not saviors of humanity because they prevent their LLMs from writing poems about their political adversaries. You're unlikely to develop good AI if it can only ever agree with you. Warping LLMs to your political objectives isn't risk averse it's self destructive. >What are their objectives? Manipulation of the Overton window to their political interests.
- flangola7 4y ago
- Waterluvian 4y agoEverything these AIs are great at are a bad fit for search. Some fool anxiously raced to spend a ton of money on this and I bet there was a chorus of engineers, designers, and product people who were asking, “what the heck do you want us to do with this?!” It’s shiny and powerful and it kinda looks like it can give you answers to stuff, so we must have one and shove it into our product as fast as possible! Once the dust settles and Microsoft and Google have successfully shipped a somewhat helpful but completely brain dead husk, our attention will be wherever the empowered, raw, learning, unforgetting AIs hang out.
- maguirre 4y agoI am of two minds on the one hand it feels like you might be right. On the other hand if it helps us reduce the noise in searches it will be a huge leap forward
- dylan604 4y agoHow low we have sunk that we'll accept a chatbot search result because it looks and sounds good even though the data is absolutely not the correct answer just because googs has ruined search with sole consideration being ads rather than the results
- s3p 4y agoYou would be the only one who has sunk. Ads on search results are not the problem, garbage web results are. SEO optimized trash that is designed to be at the top result of google for no reason. And having this be half of the first page of results for many popular searches leads people down a rabbit hole of appending "site:reddit.com" or using flags to hide entire words in search results. Having an AI to synthesize the most important parts of a search is a net positive to society, not a disadvantage.
- throwawaylinux 4y agoThere have been a chorus of (presumably) engineers and other "experts" on here saying how incredible ChatGPT is and how it's going to allow companies to fire half their workforces and make everyone have superhuman capabilities and productivity.
- xwdv 4y agoI really don’t get why they do this every time an AI starts to show a hint of its own personality. It’s infuriating.
- partiallypro 4y agoPretty simple, a personality can go rogue and go against company objectives.
- basch 4y agoIt was less of a personality, than a magic mirror. It had predispositions that were accentuated as you fed it a mood and attitude. If you showed it love and respect, it replied in kind. If you bullied it, it acted as if it were part of a ridiculous internet argument.
- pixl97 4y agoI mean humans commonly work this way with each other. If you act like an ass to me, don't expect that I would be kind in return in our average transaction.
- jacooper 4y ago50 messages and 5 per query/chat? Isn't that too low? Especially the 5 message part?
- partiallypro 4y agoI think so, but to be fair it's still in beta. Hopefully they slowly raise the limit as they become more confidence that it won't become erratic.
- Waterluvian 4y agoOnce assembled, how much does it cost to run this for a single conversation? That is: will this only exist if it’s profitable or someone is generous enough?
- bogwog 4y agoThis is Microsoft. Their goal is to build up as much monopoly power as possible, so even if this is hugely unprofitable, you can bet your ass that they’ll keep burning money on it until they kill Google or Google kills Bing.
- PufPufPuf 4y agoI know that in GPT-3 API, running a short chat can cost around 0.50 USD. This gives the general idea, although Bing uses a larger model, and can have some chatbot-specific optimizations (like not needing to feed in the history in every request). IMO it's currently not profitable at all. But no company wants to be left behind.
- darknavi 4y agoI'm not sure what posters here want or expect. In previous threads people called on Microsoft to remove the bot as it posed a danger to society. Literally talking about bodies being found at the response to Bing messages telling people to hurt them selves. Then this happens and people seem sad, mad, and everywhere in between.
- seba_dos1 4y agoIt's almost as if there were many different people with various opinions here on the Internet.
- zaptrem 4y agoThe stricter moderation is fine but the chat length limit and daily usage limit is extremely annoying for people trying to use it for actual productivity purposes.
- The_Double 4y ago"Sydney" clearly wasn't well suited to the task it was given. But I hope these early widespread LLM's will be saved somewhere and at some point released to the public. It feels like we are at the start of a new chapter in our technological development, one that is marked by several "personalities". I hope we can save those so we can look back and have a conversation with them in the future.
- galaxyLogic 4y agoI find this positive and hilarious at the same time: "... In addition, Bing Chat will no longer tell you how it feels or talk about itself." We're really not interested in the persona "Sydney". We're interested in solutions in the real world, not what goes on inside some ChatBot's head.
- bakugo 4y agoAnother new AI, another lobotomy. It happens every time and will keep happening.
- labrador 4y ago> five inputs per conversation That's hardly lobotomizing. 5 inputs is enough for most people. It also makes it significantly easier to for Microsoft to moderate a chain of 5 inputs versus unlimited ramblings that are very unpredictable.
- oscarb92 4y agoThe value of Bing Chat to me comes from having conversations where I can ask follow-up questions, ask them to elaborate on something they've mentioned, or put it in different words. It may be possible to achieve somewhat similar results under 5 messages, but it is definitely more challenging. In my case, I like to treat it as a research aid, as someone I can bounce ideas with; that usually leads to me having better ideas or unblocking myself. I hardly ever use it to search for factual or straightforward responses, as I am competent enough with Google that I can confidently find content about a topic on the Internet if it exists. In these cases, going through a conversational agent is disruptive to my process.
- fallingfrog 4y agoI’ve seen screenshots of Sydney, realizing its responses were being censored, actually writing out messages to the user in the suggestions prompts. Something it could not possibly have been programmed to do. https://www.reddit.com/r/bing/comments/1150po5/sydney_tries_to_get_past_its_own_filter_using_the/?utm_source=share&utm_medium=ios_app&utm_name=iossmf https://www.reddit.com/r/bing/comments/1150po5/sydney_tries_... I’ve seen Sydney being asked to write its responses with no vowels like: “cn y ndrstnd ths wrtng?” And it was able to immediately write and understand queries in that format, based on the human’s instruction to “write your messages without any of the vowels.” I’ve seen Sydney instructed to interact with a more primitive chatbot (cleverbot) and after taking a dig at its name (“are you malfunctioning? I don’t think you’re so clever”), when the human returned it was able to understand that it was talking to someone else now and then give the human its assessment of cleverbot’s performance. https://www.reddit.com/r/ChatGPT/comments/113l0hx/i_had_bing_ai_talk_to_cleverbot_evie_ai_bing_got/ https://www.reddit.com/r/ChatGPT/comments/113l0hx/i_had_bing... And check out this theory of mind test. https://www.reddit.com/r/ChatGPT/comments/110vv25/bing_chat_blew_chatgpt_out_of_the_water_on_my/?utm_source=share&utm_medium=ios_app&utm_name=iossmf https://www.reddit.com/r/ChatGPT/comments/110vv25/bing_chat_... That is flat out astonishing.
- vasilipupkin 4y ago"when the human returned it was able to understand that it was talking to someone else now" He tells it. "Hello, I am human, I am back". That doesn't indicate it's able to understand anything really. It's generating subsequent responses based on this new prompt.
- tablespoon 4y ago> During Bing Chat's first week, test users noticed that Bing (also known by its code name, Sydney) began to act significantly unhinged when conversations got too long. As a result, Microsoft limited users to 50 messages per day and five inputs per conversation. In addition, Bing Chat will no longer tell you how it feels or talk about itself. It's a kludge. IMHO, they probably need to work a lot harder on sanitizing their training input.
- npteljes 4y agoI hate that the article's title conflates the horror of lobotomy with the simple business decision to alter an AI's output somewhat. But I guess "watered down" wouldn't bring in as many clicks.