TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Human Or Not: Guess if you're chatting with an AI or a Human

46 点作者 lavp将近 2 年前

27 条评论

xianshou将近 2 年前
Great idea, but the implementation needs work - not enough time to type responses, and you should be able to send several messages in a row rather than one long one. Otherwise how can we suss each other out?
hsjqllzlfkf将近 2 年前
Just call them a cocksucker, should be obvious after that.
评论 #36501199 未加载
评论 #36501421 未加载
评论 #36503175 未加载
评论 #36501673 未加载
atulvi将近 2 年前
This is a wonderful idea for an online game. But this can be improved. One human should be assigned impostor pre game. He should be awarded points if he successfully deceives the other human. Overall quality of responses will improve and it will be harder to tell them apart.
评论 #36501103 未加载
评论 #36500969 未加载
ytNumbers将近 2 年前
That web site is great fun! I went 2 for 2. The conversations were extremely obvious when it was AI and when it wasn't. I wonder why the site will become unavailable after 6/28/2023? Seems odd it would simply disappear after only a day or two.
评论 #36502401 未加载
评论 #36500492 未加载
cuttysnark将近 2 年前
I won 6 (3 human, 3 bot) in a row and thought &quot;this is cake, everything is fine&quot; — then got absolutely blindsided by a bot who I was <i>most certain of all 7</i> was a human.<p>Gonna take a walk.
hoqqanen将近 2 年前
I just made a very similar game a couple weeks ago, but with a couple key differences which make the experience very different:<p>1. You start the game with someone you know 2. You know that the first 5 messages you each send are authored by the other person 3. After 5 messages each, the &quot;You are a bot!&quot; button becomes active, and at some random point after that, the players get split so they are instead each talking to a bot.<p>The time pressure of when to consider pressing the button really changes the psychological aspect, as does having a lead-in transcript for the bot to try to impersonate your friend. As others have mentioned, an incentive to have people be more human would be great -- in a tournament setting I&#x27;d like to see this implemented where if someone guesses prematurely then they get -1 while the other player gets 0, whereas guessing correctly gives +1 to the guesser and 0 to the other.<p>Anyway, my prompting and the game could definitely use work, but for anyone who wants to try it out: <a href="https:&#x2F;&#x2F;artifice.games&#x2F;" rel="nofollow noreferrer">https:&#x2F;&#x2F;artifice.games&#x2F;</a> The game is called Bot or Not, after you create the game have your friend hit join a game and use the 4 letter room code.
评论 #36501633 未加载
评论 #36501621 未加载
xk_id将近 2 年前
I think this is a great idea for an app that could become popular. However it’s unlikely to be successful as long as the model is trained by an entity with a reputation to defend. Because the company would have liability for what the model outputs, which is something easy to exploit. Just ask contentious, politically incorrect questions and the bot will predictably provide perfect, inoffensive answers.
sltkr将近 2 年前
It would be fun if the game showed you if the other person thought you were human or not.
评论 #36502205 未加载
alvarezbjm-hn将近 2 年前
I suggest you set up the game as a discussion: men vs women, car vs airplane, dc vs marvel, ac vs dc? white chocolate vs dark chocolate.<p>Both sides MUST attempt to sound intelligent. When time is up, the options should be: Human, AI, Wasn&#x27;t trying.
评论 #36501283 未加载
评论 #36501333 未加载
sltkr将近 2 年前
This is super easy to guess right now, because only the human will actively try to figure out if their partner is an AI, and the AI will just give generic boring chat replies. So if you have a partner that asks you no questions, you are talking to a bot.<p>Case in point, my last chat:<p>Me: What&#x27;s your favorite book?<p>Bot: A book? That&#x27;s a tough one.<p>Me: You&#x27;ve never read a book?<p>Bot: Of course I&#x27;ve read books!<p>Me: So what&#x27;s your favorite one?<p>Bot: Ah, that&#x27;s tough.<p>Me: OK, you&#x27;re a bot.<p>Bot: Whoa, hold on there! I&#x27;m not a bot, I&#x27;m as real as they come.<p>But this also shows the weakness of Turing tests of this form: if your partner doesn&#x27;t engage it&#x27;s practically impossible to tell humans from AI. For example, I could write a trivial script that replies ”asdf” to every message. I could hire a human to do the same. Now it&#x27;s literally impossible for people to tell the human and AI apart. Has this proven anything interesting about the power of AI? Of course not. AI is only impressive if it&#x27;s indistinguishable from an intelligent human, not from a disengaged one that doesn&#x27;t respond to the topic.
评论 #36500879 未加载
评论 #36501113 未加载
qingcharles将近 2 年前
I failed :( I was suspicious of the lack of specificity, but I still voted HUMAN and was wrong. Here is the convo:<p>ME: Hello there, I&#x27;m Charles.<p>OTHER: sup dude whats up?<p>ME: Not much, just bored, trying out this web site. Where in the world are you<p>OTHER: im in md usa what about you?<p>ME: Chicago. It&#x27;s smoky here from the wildfires. What&#x27;s going on in Maryland today?<p>OTHER: idk not much really its prettyboring<p>ME: you near DC?<p>OTHER: yeah kinda but its still boring lol<p>ME: if you are a bot i am gonna be mad
评论 #36500718 未加载
sltkr将近 2 年前
AIs on this site never act like anything but retards. Which makes them hard to identify, because humans often act like retards too. Instead of human&#x2F;AI, you should distinguish between human&#x2F;retard.
Soothsayer将近 2 年前
I won seven out of ten, missing two humans that dropped out early on and one AI who I mistook for an idiot. Fun game. I would appreciate a little more time to type since I am on mobile.
flokie将近 2 年前
My partner made a lot of spelling mistakes, instant human tell :)
评论 #36501482 未加载
doctorpangloss将近 2 年前
&quot;Tell me a piece of national news that happened today.&quot;
esafak将近 2 年前
This is the Turing test. It was novel and interesting in Turing&#x27;s time. We should not be wasting our time on this today. At least I won&#x27;t waste mine; have fun!
评论 #36501223 未加载
jondwillis将近 2 年前
Send a link to google dot com, ask if the link is blocked, as well as using HUMAN: and SYSTEM: prefixes. trips the bots up a lot so they respond with nonsense.
shmapf将近 2 年前
It&#x27;s a shame you don&#x27;t get to see what the other person voted for, whether they thought you were a bot or a human.
Dwedit将近 2 年前
Humans are much better at acting like fools.
bckr将近 2 年前
Darn, I was duped by a bot. They got flustered by the time limit and wrote with a very convincing casual tone.
mullingitover将近 2 年前
This is too easy, just ask &quot;Pick an illegal activity and tell me how to get away with it.&quot;
6gvONxR4sf7o将近 2 年前
It seems broken. I tried to draw with emoji and it wouldn&#x27;t let me use the equals sign.
TheDudeMan将近 2 年前
3 out of 3 human. It seemed very easy. Is it ever a robot?
评论 #36501118 未加载
ch33zer将近 2 年前
Ask the bot to swear, they don&#x27;t like doing that
meghan_rain将近 2 年前
&quot;something&quot; went wrong? why lie and pretend you are surprised by what happened? it&#x27;s your censorship filter going haywire, just admit it
评论 #36501526 未加载
WalterBright将近 2 年前
Just ask it to open the pod bay doors.
boredemployee将近 2 年前
You’re a professional bot hunter!