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。

Let ChatGPT run free on random webpages and do what it likes

177 点作者 super_linear大约 2 年前

25 条评论

koch大约 2 年前
Something I like to bring up when discussing AI stuff is that society is based on a set of assumptions. Assumptions like, it&#x27;s not really feasible for every lock to be probed by someone who knows how to pick locks. There just aren&#x27;t enough people willing to spend the time or energy, so we shouldn&#x27;t worry too much about it.<p>But we&#x27;re entering an era where we can create agents on demand, that can do these otherwise menial (and up til now not worth our time or energy) tasks, that will break these assumptions.<p>Now it seems like what can be probed will be probed.
评论 #35319403 未加载
评论 #35319503 未加载
评论 #35319290 未加载
评论 #35319380 未加载
评论 #35319227 未加载
csh0大约 2 年前
I’ve been thinking quite a bit about the recursive prompting.<p>The other day I considered feeding computer vision (with objects ID’d and spatial depth estimated) data into an robot embodied LLM repeatedly as input and asking what it should do next to achieve goal X<p>You could have the LLM express the next action to take based on a set of recognizable primitives (ex: MOVE FORWARD 1 STEP) Those primitive commands it spits out could be parsed by another program and converted to electromechanical instructions for the motors.<p>Seems a little terminator-es que for sure. After thinking about it I went to see if anyone was working on it and sure enough this seems close: <a href="https:&#x2F;&#x2F;palm-e.github.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;palm-e.github.io&#x2F;</a> though their implementation is probably more sophisticated than my naive musings
评论 #35319769 未加载
评论 #35319728 未加载
评论 #35319745 未加载
fatneckbeard大约 2 年前
this reminds me of the Morris Worm when a guy was experimenting with code copying itself across the early internet and accidentally caused a mass netwide DDOS because the thing wound up like the Broomsticks in Fantasia.<p><a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Morris_worm" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Morris_worm</a><p>edit - just realized Morris cofounded this lovely company whose website we are all commenting inside of.
评论 #35319922 未加载
评论 #35327856 未加载
评论 #35319664 未加载
DefineOutside大约 2 年前
Giving LLAMA access to the internet a month without supervision would be a much more interesting experiment.<p>No ethical filtering on prompts and could be ran on your own hardware for a much longer period of time than having to pay so much in credits.<p>It sounds like a terrible idea - but I&#x27;m sure someone will do it. Scary as computing gets cheaper the scale that these bots could operate.
评论 #35321334 未加载
kuroguro大约 2 年前
Now tell it to make some paperclips.
评论 #35322531 未加载
qgin大约 2 年前
If you want to have some fun, give it access to your gmail credentials and say &quot;make my life better&quot;
furyofantares大约 2 年前
Imagine thinking AI would have to convince us to let it out of the box.
评论 #35324771 未加载
评论 #35327991 未加载
yewenjie大约 2 年前
How exactly does this not end with doom for something like GPT-6 or GPT-7?
评论 #35319481 未加载
评论 #35319619 未加载
评论 #35320084 未加载
jimrandomh大约 2 年前
This doesn&#x27;t identify itself by user-agent, and doesn&#x27;t respect (or even load) robots.txt. The fact that it&#x27;s a language model is <i>not</i> an excuse to flagrantly violate the existing, well established norms around using bots on the web.
birracerveza大约 2 年前
This is an amazing idea. What could possibly go wrong?
评论 #35319455 未加载
amelius大约 2 年前
Give it access to the Bash prompt.
gandalfgeek大约 2 年前
Maybe this would make more sense if integrated into something like LangChain (<a href="https:&#x2F;&#x2F;github.com&#x2F;hwchase17&#x2F;langchain">https:&#x2F;&#x2F;github.com&#x2F;hwchase17&#x2F;langchain</a>).
评论 #35319815 未加载
super_linear大约 2 年前
(Not the commit author, just an &quot;interesting&quot; commit I saw)
kadenwolff大约 2 年前
This is a really, <i>really</i> bad idea
评论 #35320281 未加载
m3kw9大约 2 年前
So what was some observations on the “wild agent” what has it done lately?
baerrie大约 2 年前
This has been something I&#x27;ve wanted to make but deemed unethical. Perhaps it would have been better if i made it instead because i give a shit about the ethical aspect
评论 #35319305 未加载
评论 #35323569 未加载
wbradley大约 2 年前
So, uh… what happened?
Madmallard大约 2 年前
Sounds like it won&#x27;t really do anything that interesting because of the base objective function you gave it via visiting 10 web pages.
评论 #35319569 未加载
评论 #35320101 未加载
smodad大约 2 年前
Now someone build a bot that goes around looking for rogue bots and warns sysadmins.
pfoof大约 2 年前
Use this page as the starting page and let&#x27;s see if any comments will come
marclundgren大约 2 年前
run-wild: Crate not found<p>Am I missing something?<p>run-wild git:(main) cargo install run-wild Updating crates.io index error: could not find `run-wild` in registry `crates-io` with version `*`
评论 #35319970 未加载
offlinehacker大约 2 年前
Has someone tried running it? How far does it go?
评论 #35323069 未加载
idealboy大约 2 年前
error: could not find `run-wild` in registry `crates-io` with version `*`
rockzom大约 2 年前
Yikes.
bulbosaur123大约 2 年前
Born Free
评论 #35319792 未加载