INTRODUCING PEAR trypear.ai !!!!!!! I will be live on twitch.tv/fryingpan tomorrow (Sunday) and introducing the entire project. If you'd like to participate in UX, coding, ideation, or simply enjoy the journey, you are welcome! TLDR: We are aiming to reduce the time from ideation to conception by building the next-gen product building environment: an open-source AI powered code editor, PearAI. If you like the vision, let me know in the comments :) ► What is this startup series? 2 best friends living in New York, @FryingPan, a software engineer who worked at multiple big tech companies, and @nang88, a software engineer who worked at a top high frequency trading firm, who just recently quit to do startups, join together in search of a way to create positive impact in this world. Our current idea? PearAI, an open-source AI powered code editor. Join us on our mission, either as a user or contributor, to empower people to build out their ideas by reducing the time from ideation to conception: trypear.ai. (PearAI Discord: discord.gg/7QMraJUsQt) Timestamps: 0:00 - Intro 0:25 - Introducing our new PearAI idea 2:16 - Coworking at my apartment in New York 2:45 - Friday night market break 3:40 - Weeknd work - First Pull Request 4:00 - Discussing pivoting ideas once again (sick in bed) 5:03 - Coding: GPT integration 5:55 - Our plan and vision with this idea: Ultimate Goated Plan 7:25 - Buying a domain name and setting up landing page 8:10 - YCombinator (investors) preparation & interview 11:30 - Conclusion... join us:)
this is exactly what I thought before they started saying "codebases". So I thought "Oh, they're using it to assist with entire codebases/files rather than just one singular file.
@@OompaLoompa74 yeah, copilot does that.. and there's github copilot workspaces now. nonetheless it's a pretty good technical project to have on a curriculum
video is really nice, but whats the value proposition for the product? its just vscode with ai integration, there are already completed projects with big teams behind it like e.g. idx. would love to hear you talk more about the actual project, what it entails, what problems it solves and why its better than idx etc. looking forward to your future videos
Amazon has an internal tool just like this, maybe u can conduct some interviews from aws employees to ask them how they use their tool, which is similiar to copilot
How long to make that perfect toilet angle? Genius UPDATE: I have 0 clue how to comment ending up under this video (it wrote this under another vid) Someone at RU-vid please explain 😅
dude, your videos are great man, I started watching indie makers 2 days ago, starting with your video, and then the yt suggestions, just connected me to other people who are trying out new cool things. Thanks bro, you inspire me. I hope I can also build great things.
The problem with AI wrappers for software engineering is that the models are simply just not good at programming. Co-pilot makes mistakes constantly working with complex things. The amount of information it needs to make a sound decision will basically be your whole product structure. Ai right now just isnt at a place where it can make reliable software cause most code its trained on is shit. All models right now have this problem they can't even solve new leetcode mediums.
brother I like your videos but theres no viable path and wayy too much competition have u considered the fact that u are doing startups cause u want to id say taje more time and when u find a good idea then do it pivoting ine industry to another shows lack of base idea
Always fall in love with both pan and nang content. As a backend developer love the idea of a AI powered code editor. At least it will help in reducing the tab switch from chatgpt to code editor 😅. I'll also loved to contribute 💛.
How is this different from Zed, where you can integrate ChatGPT or other AI Chatbots with tokens; also Zed is open source, in Rust, and built by the founders of Atom?
Wish you guys the best of luck, however I have a really hard time seeing how 1) this is unique at all 2) how you're gonna compete against multi million dollar backed alternatives
please make a website or platform that connects "larger" RU-vidrs with sponsors directly, so don't have to go through agencies that take an unknown percentage
Tbh, a lot of these ideas are going to get done a lot quicker and a lot better by openai, Microsoft and Google. I know its hard to hear buts it's the truth. A lot of people have already correctly pointed out that this is similar to copilot. This is good as a side project for fun and for your cv, but the real advantage and potential in AI/deep learning applications lies in going deep into a niche or specific industry. What are some niche problems that you guys have that isn't hot right now, you want to position yourself to catch the wave, not to chase the wave that has already come. But of course the American market is so huge that it's possible to even capture such a small part of it and still make some money. All the best.
Competition is always good, theres a lot of products out there that suck, are completely overpriced or dont fit consumer need. Niche usecases help but always think how it can be used to spring board into a wider audience
@dandogamer yes, niche uses provides that competitive edge. Bigger companies tend to only focus on mainstream use cases because that is the only thing thar can earn them satisfiable profit
I like these but there are already a lot of tools for this market and the big guys are already somehow perfected this to a high level. So it will be a challenge for you guys but goodluck
I can see this going either really good or really bad, I think you can learn a bit about the business strategy used by framework (the hardware device company) in how they pivot their business strategy around how mature the market is. Theres an interview on yt on how the CEO of framework utilized this existing market to their advantage. All the best.
i can't believe i got spoiled by watching the interview on nang's irl channel beforehand XD, it's fine tho, it gave me a little more understanding about the project vision in general
how will you make money if it's open source? please dont do ads lol. as others commented, you are basically building what other AI editor company is already building. they can easily add the feature that many people request but you have to build it from ground up again just to catch up with them. they also have a bigger team than you. it seems like you are just building something that fits your needs exactly instead of addressing the wider market.
instead of just a text you can add a video to the website that explains every thing in it with a good edit and for advertizing you should use short form videos reels tiktok shorts and make a list for all companies you want to work with then send them emails cuz actually the way that you are using in the video (walking from place to another ) is a slow way all what i said was my opinion so you can consider it or ignore it but any way keep guys going and dont give up
Dont yoy think that the large firms are already working on this i mean this is the ultimate goal of ai , to reduce time so how do you plan to compete with let's say chatgpt comes up with a code editor of their own ?
I'm going to be a junior cs student in the fall, I just took a look at your slides and github and this project looks so cool!! I would apply but I fear I don't have the experience you are looking for lol. I wish you the best of luck on your start up journey :)
Everyone thinks they can make a product better than any existing ones in the market until they realize how dominant the existing ones already are in the market.
it would be cool if you guys could make an AI finance tool that links to bank accounts and predicts how much you'll have in expenses next month, how you can reduce them, how you can reduce tax and write off expenses etc
@crafter505 you can use bank apis that only have access to their transactions. Plus higher barrier to entry which is a plus and you can start by making it for yourself which wouldn't need to be as secure
@@EclipsyChannel Wrong it's $0.35 USD, the ''official'' dollar that you see on online converters are the dollar that the banks offers, but that one isn't used anymore, for online payments we use the ''Card'' Dollar which costs $ 1.448,80 Pesos, then there is the Blue Dollar which is the most used irl, which costs $1.120 Pesos, everything got fucked in Argentina when it comes to dollars, the average salary rounds $750USD but rent is around 250-400USD(Out of the centre), then food rounds 200-250USD too, besides that RU-vid took +750 Pesos from my wallet because of taxes instead of the 500 Donated, which is nothing to support financially, just to express community support.
I'm literally a dev in my 30s and i cant imagine even coming up with an idea and trying to establish a startup. Massive props to you sigma grindsetters
Good luck man. Im working on my first SaaS app, as a solo dev and it is quite a lot of work. I needed compute for it & EDA, so I just decided to use AWS. Now I have to maintain 5 different repos with their own github actions workflows that execute their own CI/CD pipelines & push images to lambda functions. Perhaps this is a skill issue in my design, but i don't know how solo devs build things like this without going insane.
