| Author | ChatGPT is blowing my mind |
|---|
"Write me a selenium python script to start a hunt in lordswm.com"
from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
# Replace the placeholders with your username and password
USERNAME = "your_username_here"
PASSWORD = "your_password_here"
# Create a new instance of the Firefox driver
driver = webdriver.Firefox()
# Navigate to the login page
driver.get("https://www.lordswm.com/login.php";)
# Find the username and password input fields and enter the credentials
username_field = driver.find_element_by_name("login")
password_field = driver.find_element_by_name("password")
username_field.send_keys(USERNAME)
password_field.send_keys(PASSWORD)
password_field.submit()
# Wait for the page to load and navigate to the hunting page
hunt_page_link = WebDriverWait(driver, 10).until(EC.presence_of_element_located((By.XPATH, "//a[@href='/hunt.php']")))
hunt_page_link.click()
# Find the Start Hunting button and click it
start_hunt_button = WebDriverWait(driver, 10).until(EC.presence_of_element_located((By.XPATH, "//input[@value='Start Hunting']")))
start_hunt_button.click()
# Wait for the hunting to start
WebDriverWait(driver, 60).until(EC.presence_of_element_located((By.XPATH, "//div[contains(text(), 'You are hunting now!')]")))
# Close the browser window
driver.quit() |
In this script, we're using the webdriver module from Selenium to create a new instance of the Firefox driver, navigate to the login page, find the username and password input fields, enter our credentials, and submit the login form. Then, we wait for the page to load and navigate to the hunting page by clicking on the corresponding link.
Next, we find the Start Hunting button and click it. We then wait for the hunting to start by waiting for the presence of a message that says "You are hunting now!" to appear on the page.
Finally, if everything went well, we close the browser window using the quit() method.
Note: You will need to install the selenium module and have the geckodriver executable in your system PATH for this script to work. |
It's insane to me that the 100s of hours I spent writing different scripts can now all be automated within minutes.
The future is now and I'm absolutely loving it. The possibilities are endless. Truly.
What do you guys think. |
Btw this truly is an off-game thread, the script above was purely as a relatable demonstration of what it's capable of.
Just starting using chatGPT 2 days back and so I'm exploring the edges of what code I don't have to write myself anymore. |
Do note to all players that all scripts have to be vetted by admin before using. If not, you might get jailed/finned/blocked. Automatic scrips are NOT allowed.
for ProZyk:
You ought to be careful, you know what happened in the past. |
Very OFF-GAME now.
I just got chatgpt to convert my vimscript commands into lua for me.
I commit to never learn Lua ahaha.
Now I'm getting it to diagnose why my gopls is so slow, and wowowow.
I can't use chatGPT for work yet, but man am I blowing through all my pending personal projects in record time. |
Do you really think that this is the right place for such discussion when we all know what happened in the past as Arcanide said?
Rubbing on the shepherds stick was never a good idea. |
| What exactly had happen in the past? |
for - Concord -:
prozyk got banned for illegal script |
| Seriously, just delete my first post please. There's so much more to discuss there than just it's impacts on scripts. |
I get your point, but funnily enough, I disagree.
I'd argue that chatGPT's biggest benefit on programming right now is making boilerplate scripts easier to implement without searching through confusing documentation/tutorials. |
That was the most annoying part of programming to me. So I'm glad it's the one getting fixed first.
You can get it to debug anything you want. Send it an error message and possibly some context it will construct the story around it. It's akin to having a sit-down with the author of a library, except he's being nice this time.
I'm also having success in using it as a teacher. It's teaching me Spark now, and I love being able to question it along any tangent and it having reasonable answers to them. |
| Reasonable in theory, sure. |
| We found at work that it tends to make things up, including entire scientific papers to reference that dont even exist |
| The script example you shared is pretty basic one, I doubt it can write very complex stuff, for example let’s say auto enroll, I doubt it can write that |
I’ve tried to do Development with it, you say hey Chat gpt make me a login page, it’ll give you one, you say hey make a css and make it more beautiful it’ll do that, you say hey write authentication code it’ll do that
But I somehow still didn’t find it as good as writing and doing things on own, it lacks the human touch, and I guess that’ll always happen, hence not it’s biggest fan, though sure it’s great at writing descriptions Etc
Just my koi onion |
| ChatGpt is good for lazy people or when you need a draft of 80-90% accuracy, so you can start with it and improve the outcome until 95%+. You save a lot of time. |
| Cool stuff. I saw someone was asking it question about LWM, and it was confidently giving fake information, like how to exhaust griffin’s retaliation with a fairy |
for ProZyk:
It sounds useful but perhaps you may be overestimating it. I just have this feeling.
Myself, I'm unfamiliar with chatgpt but it's uncanny how it was introduced to me very recently. I was having a conversation with a medical student who was talking about this software when making a point about how the field of diagnostic radiology was destined to become obsolete in the near future due to AI being able to recognize abnormal patterns in imaging (CXR results, MRI results, CT, PET, etc.). I know it would sound plausible to a layman how that could be true, but I was dumbfounded to hear about it from a senior medical student, knowing the complexity of imaging interpretation, the nuanced distinctions in every result, and various background considerations that go into making an irrefutable diagnosis.
I'm no Luddite and I sincerely find that we are so fortunate to have all sorts of tech in our practice, so grateful indeed. But that said, I think no matter where you reach with the AI, humans will always feel most satisfied with the "human touch" in every little thing. It's the same way whether we talk about something mundane such as the use of automatic cars or something more serious like robotic CABG (basically heart surgery using robots), people will always need a human to cross-check and give them assurances about the quality of the performance.
For clarification, I do think AI is never going to be competent enough to handle the jobs in the examples I listed with 100% independence, but I leave that to your own imagination and assessment. I wanted to make a point about this inherent psychological bias most of us have, and it isn't really possible to easily overcome. |
It's the same way whether we talk about something mundane such as the use of automatic cars or something more serious like robotic CABG (basically heart surgery using robots), people will always need a human to cross-check and give them assurances about the quality of the performance.
Your examples are interesting - of course, AI will almost always make mistakes, but it's not about being perfect - it's about being better than humans, which is certainly achievable.
The biggest issue with AI is assigning blame. Do we want to make the manufacturer of an AI-powered car responsible for accidents where the "AI" was at fault? If so, then no manufacturer would invest it in - it just wouldn't be fiscally practical. |
