I Used AI To Scrape The Web & Write PDF Reports
HTML-код
- Опубликовано: 16 авг 2023
- Let's dive into GPT Researcher to see what it can actually do and whether it is useful. GPT Researcher is an autonomous agent designed for comprehensive online research on a variety of tasks.
Want more content?
🔐 Copy my AI tools & workflows: bit.ly/data-alchemy
💰 Kick-start your data freelance career: www.datalumina.io/data-freela...
🔗 Links
github.com/assafelovic/gpt-re...
app.tavily.com/
🕷️ Debugging
conda install -c conda-forge uvicorn
conda install -c conda-forge weasyprint
pip install --force-reinstall httpcore==0.15
Chrome driver: Apple System Settings → Privacy & Security → Allow
👋🏻 About Me
Hey there, my name is @daveebbelaar and I work as a freelance data scientist and run a company called Datalumina. You've stumbled upon my RUclips channel, where I give away all my secrets when it comes to working with data. If you want to learn more about what I do, then head over to www.datalumina.io/ 
Наука
🔐Copy my AI tools & workflows: bit.ly/data-alchemy
Wow, your video is incredibly informative! One key takeaway I found from your video is how you effectively navigated through the challenges and bugs that can come up when using open-source tools. Your troubleshooting approach is commendable. Can you also share some insights or tips on how you plan to leverage the research report generation capability of GPT researcher in your future content creation or projects?
Thanks for sharing!
This is amazing!!!
Hi Dave
great video! I have a question: lets say if you have 100 of pdfs . and I want to create a research report taking information from all the pdfs with just 1 question that user asks. How do you think we can do it. Please can you make a video out of it
great video - thank you
nice video! i wonder, is there a way to run this completely locally, including the AI? say, WizardLM 70b. should be an insightful project to show us how to do this.
why?..."just b/c we can so we can understand it." love that comment😁
Hi Dave, great video on the process. Did you try to limit the access to retrieve information from trusted websites? In research the source is important.
How to use it with open source models ? maybe using llm studio to use on cpu, or any other local models on gpu
Suggest some good free alternatives of open ai
Constantly having errors related to chrome timing out etc
Nice UI and but the report this tool generates is of low quality unfortunately ... thanks for sharing
Hi Dave, I really like your video, but I have a problem. I live in America and there is a lot of squawking about AI-produced content to the point that people are using AI detection for content and some organizations are disallowing the use of AI-produced content.
Now having seen what you did I admit I am uncomfortable. I would never have such a great report, produced and what would I produce as AI-generated, I didn't write any of it how can I take credit? I could take credit, but then someone runs an AI detection on the report and it comes back 100% AI written. That's not good after I just said I wrote it.
Hollywood screenwriters are terrified that this report generated can be trained to write screenplays and do it quite well.
So, what I am saying it is great as an assistant to write a research paper,l but it cannot be the sole content.
You explain how you would like to produce content automatically and place it out there with it have any human content, today people not digging that idea. Imagine, an AI agent producing content on a subject uncontrolled by a human giving humans information to follow, that doesn't sound right. The idea you are putting an AI agent out in the world with instructions to do whatever. I'm not saying it's bad, but you are praising its use that way. It's an ethics thing that is going on in the US, perhaps where you are you all trust each other enough to not have a worry about the abuse and misuse of an AI agent. Good video, and a good presentation. This is just a conversation and I'd like your opinion when you get some time. take it easy.
The process you described is called evolution. There are countless jobs and practices that have been destroyed by technology. With technology I don't mean electronic stuff, but the original greek term instead which broad. Think about the world of communication: in ancient times, there was no way to communicate between two human beings if not by traveling the entire distance that separated them. The demand, especially in war scenarios, created the role of the messenger. That role has accountability and responsibility, but you also took credit for the outstanding actions if you were well articulated in communicating the message and quick to reach the destination. Just like writing a paper or a film script, you are both responsible and praised. Then, carrier pigeons burst into the scene to cover such distance. Both people and animals were unreliable because they made mistakes and were quite slow. Everything changed with the telegraph, and the later evolutions (like the protocols behind internet), fast and 99% reliable. Every role that had to do with transporting information decayed. People hardly send physical mails and postmen are rare. When it comes to ethics, the bad intentions have always been there: voluntarily changing the message to communicate because the messanger was corrupted, man in the middle hacks in the internet age, but it doesn't mean that the technology is not right. Good people will generate content with AI, parse it, modify it if necessary and take credit as reviewers (and rightfully so). Bad people are going to use it for propaganda, fake news, fraud etc.
It's the human nature that we have to question, not the technology. Evolution is unstoppable. Of course the discussion will be very different once we reach fully autonomous AGI, but that's another story.
You are supposed to disclose that it was AI generated as per the terms.
@@danjoseph5707
Hi Dave. thanks for responded to my comment, however that was 3 months ago, when I virtually knew nothing. I don't think I have to discuss when my writing is AI assisted today.
As you know, everyone uses Grammerly to improve their writing, so it isn't necessary to confess your errorless writiing was due to an AI driven error correction program.
You are correct however about terms where you necessarily have to admit whatever you used to assist your writing has AI software that , in part, has assisted you in writing what you wrote, it is, indeed, a serious matter.
A person makes a choice whether they want to be part of a stringent identification program. Be Well.
@@chrismachabee3128 Grammarly is not the same as OpenAI GPT. Improving grammar is not going to pollute future data too much. Generating full scientific research without disclosing Ai wrote it is unethical. Why is that so hard to comprehend?
@@danjoseph5707
You're lucky I'm tired, I don't feel like fighting. Look dude your the one that brought stringent AI identification. I only agreed and raised it a bit because Amazon requires everything to be identified evenif it is a simply program like Grammerly. But you, you're brilliant and high and mighty talking about well you know Grammerly isn't the same as ChatGPT! Hahahahaahaha...No, really, wow you one bright sum B to know I didn't know that.
Look, when I finish this little note from you, I don't want to hear from you again, understand? I don't like smart azz snobs who think they are too cool for school, OK?
So, do me a favor, you go that way and I'll go in the opposite direction. Man, Biden voters are the worst. Just try and have conversation and their lording all over you. You know what, I think i won't return to this guy's video site.
I been seeing a guy who has a video site where he has a little sea creatre he feeds little crabs and fish to, This creature spits at the object to soften it up to eat. You remind me of that sea creature.
In the famous words of Elon Musk. "GFY"