You should try Magnet for this! (www.magnet.run)
https://www.loom.com/share/ceabf385871e45b88fcb512b6de6b684
295 karma · joined October 2, 2010
Open Product Curriculum :https://github.com/nicolaerusan/product-curriculum
Blogging at: https://medium.com/clay-labs https://medium.com/@nicolaerusan
-----
Magnet - The AI workspace for agentic coding
------
You can think of Magnet as your workspace for collaborating with your human & AI agent team mates. We let you quickly spin up Claude Code sandboxes for every issue, and we're also thinking about how AI can be more of a thought partner in building high-quality software.
We're thinking about this problem space more broadly than just trying to be a GUI for Claude Code (though that's already a great starting point).
These are a few of the themes we think about:
- How can we use AI to help you think critically about the features you're prioritizing and what to build next?
- How can we always assemble and provide exactly the right context for every issue you're working on?
- What are the best patterns for collaborating with your human & AI teammates, to ship the highest quality code?
- How can you best specify exactly what you want, and verify that it's what you hoped for?
Would love for y'all to try it, and I'll post a video of me building a product with Magnet a little later here - the tool's getting really fun to use!
We're also very open to feedback and try to incorporate learnings quickly! I spent a large part of this weekend using Magnet to fix most of the issues someone we onboarded Friday brought up
Given you decided to start self-hosted, are you planning on a cloud version in the next while too?
I'm curious also who you think is the right fit for this right now in terms of initial users
We're just starting to share the alpha of Magnet (https://www.magnet.run) a bit more publicly and we thought the community here might find it useful.
Magnet is free to use if you have your own OpenAI API key so feel free to give it a spin. If you pay for the Pro plan you get access to our GPT-4 based coding assistant.
The aim of this first version of Magnet is to make it easy to use ChatGPT in the context of your own codebase. It’s a desktop application that runs for MacOS (Windows and web coming soon).
If you find yourself copying & pasting code into ChatGPT from time-to-time than Magnet is for you - it makes it much easier to give the LLM context on the relevant code files from your codebase.
The product is early, and there's some exciting stuff coming up, but even this early version has been feeling useful and we've been using it to ship new features and solve bugs, so we wanted to ship it early
I'm onboarding folks, and if you have Algolia, Gitbook, Readme or most common docs platform it takes minutes to set up.
It also integrates with your Slack and Discord (both in terms of letting people ask questions and get responses from the assistant in those tools, and marking conversations you want the assistant to know about for the future).
Would love more folks to try it out if anyone's interested, and think their product would benefit from it!
It's pretty wild to see how quickly the space of Generative AI Media is coming along.
I started a newsletter on the topic, called The Art of Intelligence (GPT-3 came up with the name) with the first post going out last Friday on the topic of how far are we from AI generated videos, and simulated worlds like the Holodeck given the rapid progress of these visual A.I. Thought y'all might find it interesting: https://artofintelligence.substack.com/p/dall-e-stable-diffu...
This type of progress also reminds me of a really lovely publication from 2017 in Distill.pub, on the topic of these A.I. enabled creation tools - I think y'all would enjoy seeing what folks were thinking even then: https://distill.pub/2017/aia/
Just built a service I thought y'all might find interesting.
Rather than using HTML structure to scrape, I'm using GPT-3 to actually understand the text and pull information out.
You just ask for things you want from a site, for example a news article about fundraising like this: "The name of the company that raised funding" "Founders or CEO of the company" "URL of the company that raised funding" "A short description of the company" "A list of investors"
And get back the needed data in a JSON object! It's been pretty fun to play with and I thought some of you might find interesting use cases for it.
I think it would also be very effective at parsing PDFs (same as text).
I built a little-side project that tries to use GPT-3 to find and describe novel business ideas at the intersection of two existing projects. Was a fun little experiment to build so wanted to share it here :)
I wrote a longer piece with background and context on the project here: https://www.notion.so/Possible-Intro-Blog-Post-b4a588a48a034...
For VS Code Live Share I kept finding myself opening up Zoom and then in parallel trying to get Live Share running, which also feels somewhat finicky at times. The GitDuck experience felt a lot more complete by integrating at a different level. It also felt like it could eventually be a more suitable experience for things we often do in interviews like try to do coding interviews by combining tools like CoderPad & Zoom - though CoderPad has the nice side effect of preservice links to the interviews themselves.
Learn More About Life @ Clay Here: http://bit.ly/3b5bVX7 Want to help give the power of programming to an order of magnitude more people? Clay is the spreadsheet that fills itself - the first live-spreadsheet that brings together the best parts of spreadsheets, coding & simple automation. There are only ~20 million developers in the world, but over 1 billion spreadsheet users - come help us build the bridge that lets that much larger group tap into “programming” superpowers.
We’ve raised a Series A from the top investors in the world, and are a small, tight-knit group that is now building out our core-engineering and product teams in NYC. If you want to design and implement novel types of interfaces, and happen to be interested in Brett Victor, no/low-code tools and the history of computing - let’s chat!
Tech Stack: NodeJS, React, Redux, Typescript, PostgreSQL, Redis, Elastic Search, AWS, Serverless & Lambda, Terraform & more...
Open Roles:
- Software Engineer https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
Or write us an email at jobs@clay.run
The Spreadsheet That Fills Itself
Learn More About Life @ Clay Here: http://bit.ly/3b5bVX7 Want to help give the power of programming to an order of magnitude more people? Clay is the spreadsheet that fills itself - the first live-spreadsheet that brings together the best parts of spreadsheets, coding & simple automation. There are only ~20 million developers in the world, but over 1 billion spreadsheet users - come help us build the bridge that lets that much larger group tap into “programming” superpowers.
We’ve raised a Series A from the top investors in the world, and are a small, tight-knit group that is now building out our core-engineering and product teams in NYC. If you want to design and implement novel types of interfaces, and happen to be interested in Brett Victor, no/low-code tools and the history of computing - let’s chat!
Tech Stack: NodeJS, React, Redux, Typescript, PostgreSQL, Redis, Elastic Search, AWS, Serverless & Lambda, Terraform & more...
Open Roles:
- Designer https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786...
- Software Engineer https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
Or write us an email at jobs@clay.run
There are only ~20 million developers in the world, but over 1 billion spreadsheet users - come help us build the bridge that lets that much larger group tap into “programming” superpowers. We’ve raised a Series A from the top investors in the world, and are a small, tight-knit group that is now building out our core-engineering and product teams in NYC. If you want to design and implement novel types of interfaces, and happen to be interested in Brett Victor, no/low-code tools and the history of computing - let’s chat!
Tech Stack: NodeJS, React, Redux, Typescript, PostgreSQL, Redis, Elastic Search, AWS, Serverless & Lambda, Terraform & more...
Open Roles:
- Designer Who Can Code: https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786...
- Software Engineer: https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
- Forward-Deployed Engineer: https://jobs.lever.co/clay/3372897a-e5cf-4408-802d-f904f3ebc...
- Software Engineer Intern: https://jobs.lever.co/clay/3f27eb6d-2374-48c6-a844-fa6b86fa3...
Or write us an email at jobs@clay.run
Learn More About Life @ Clay Here: http://bit.ly/3b5bVX7 Want to help give the power of programming to an order of magnitude more people? Clay is the spreadsheet that fills itself - the first live-spreadsheet that brings together the best parts of spreadsheets, coding & simple automation.
There are only ~20 million developers in the world, but over 1 billion spreadsheet users - come help us build the bridge that lets that much larger group tap into “programming” superpowers. We’ve raised a Series A from the top investors in the world, and are a small, tight-knit group that is now building out our core-engineering and product teams in NYC. If you want to design and implement novel types of interfaces, and happen to be interested in Brett Victor, no/low-code tools and the history of computing - let’s chat!
Tech Stack: NodeJS, React, Redux, Typescript, PostgreSQL, Redis, Elastic Search, AWS, Serverless & Lambda, Terraform & more...
Open Roles:
- Software Engineer: https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689....
- Designer Who Can Code: https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786....
Or write us an email at jobs@clay.run
Want to help give the power of programming to an order of magnitude more people? Clay is the spreadsheet that fills itself - the first live-spreadsheet that brings together the best parts of spreadsheets, coding & simple automation.
There are only ~20 million developers in the world, but over 1 billion spreadsheet users - come help us build the bridge that lets that much larger group tap into “programming” superpowers. We’ve raised a Series A from the top investors in the world, and are a small, tight-knit group that is now building out our core-engineering and product teams in NYC. If you want to design and implement novel types of interfaces, and happen to be interested in Brett Victor, no/low-code tools and the history of computing - let’s chat!
Tech Stack: NodeJS, React, Redux, Typescript, PostgreSQL, Redis, Elastic Search, AWS, Serverless & Lambda, Terraform & more...
Open Roles:
- Software Engineer: https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
- Designer Who Can Code: https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786...
Or write us an email at jobs@clay.run
We're backed by top-tier investors including First Round Capital, Boldstart Ventures, Box Group, and an all-star group of angels.
Open Roles:
- Software Engineer: https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
- Designer Who Can Code: https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786...
Clay is a new type of tool that brings together the best parts of spreadsheets, coding & simple automation. Quickly connect your apps and code into automated workflows, build useful tools, enrich data sets and more! Come help us solve interesting UX & programming challenges, and reimagine how people build software.
We're backed by top-tier investors including First Round Capital, Boldstart Ventures, Box Group, and an all-star group of angels.
Open Roles:
- Software Engineer: https://jobs.lever.co/clay/d1cb19b1-88ab-4bea-9c78-6ac10c689...
- Designer Who Can Code: https://jobs.lever.co/clay/456b9e54-99be-4289-9529-6fca06786...
Cool project :) I've been experimenting with building similar little tools for myself using our product Clay: http://clay.run (you can think of it as essentially an integrated version of some of the tools you're using like IFTTT & Spreadsheets), that will hopefully eventually make building tools like this a lot simpler. We're early on and the onboarding isn't great right now (improved one coming soon), but would love any feedback you might have!
Will Autonomous Cars & Mobile Gyms Help With America's Obesity Epidemic: https://medium.com/@nicolaerusan/will-autonomous-cars-mobile...
Def, the nutritional issues are still there & so are the issues around communal activity - but I think for a lot of people I know who have long commutes to work, the opportunity to exercise in transit would be compelling.
Makes a lot of sense to just have everything managed and billed under one umbrella. What aspects do you see as the main benefits of the service? Do you have any services you are particularly interested in bringing on board?
Also, design is on point - great job team!
Let's consider what some of the things that would need to happen are to address the problems you bring up:
Cheap, virtually free sources of energy would need to play a part. To solve the issues of airports, maybe it's that we have personal flying crafts capable of vertical takeoff instead of relying on centralized airports. Maybe there's a way to get picked up by an airplane at your house (not saying I have any idea what that might look like, but it gets you to start imagining approaches around the constraints). What would possible solutions to the baggage issue you brought up be?
This thought exercise prompts is meant to prompt ideas around what things could look like in the extreme, so that we can then consider what might need to happen to get to something like that. For sure there's lots of hurdles, and some of them may actually be fundamental limits - though I'm not sure that such limits exist here.
Last point I'll make is that economic assumptions do change over time substantially. Until the industrial revolution in the 18th century, economics looked very different than it does today - GDP per capita did not increase as any increase in GDP was met by a proportional increase in population (https://en.wikipedia.org/wiki/Malthusian_trap). The assumptions we have about how our economy functions today could dramatically shift over time as technological advances shift the dynamics of the system. Abundant food, energy & energy storage, along with automated skilled robotic labor, could challenge many of our current economic ideas.
Not sure how things shake out as these transportation trends evolve: do you still have a tendency to crowded areas, or do you actually end up getting more diffusion. Suburbs were the city planner's utopia for a long time, but traffic and feelings of lack of community ended up making them less appealing than originally imagined. Maybe the suburb approach becomes more compelling again if problems of traffic & lack of community are addressed.
Conceptual debt is product design debt like you're suggesting. In particular, I think of it as a subset of product design debt that has to do with domain modeling as you suggest vs. having the write domain models but having poor user flows around those core concepts.
Unlike user flows, concepts are also reflected in the API & Object Models in your codebase so those may be trickier to change than the user flows that revolve around them.
And to your point too about first thinking about the conceptual model, agree that you should spend at least a bit of time before doing anything else. Thinking through your modeling is something that doesn't take that much time and has such huge payoffs in terms of avoiding pitfalls. But often times it's just funner to start writing code so people do that.
It's always saddening to see programmers just dive in and start coding without planning / thinking about the problem first.
But the concepts are not only in the product design, but also reflected in the software system architecture, since that will reflect the design.
One weird thing that happens is that developers insist that there is some sort of technical constraint, and then the product design reflects the software system architecture, rather than the software system architecture reflecting a product design well suited to users' mental models.
I know what you mean about changing underlying data structures too. There are ways to fix the issues after the fact. E.g. wrapping existing concepts in other ones, and then gracefully deprecating the old ones. But, it definitely creates all sort of weird tensions with customers that are using the old concepts.
I've been adopting this rule of thumb that the first pass at a new product I can just dive in and treat it as a throwaway, since I probably don't understand the domain well enough to develop an ideal conceptual model anyways. The key here being to treat it as a throwaway and then return and do a rewrite from scratch once I have a better understanding of the problem.