open source AI will save the world // Host @ToolUsePodcast // AI and Engineering Lead @ BoxOne Ventures
- Tweets 3,380
- Following 153
- Followers 3,778
- Likes 9,064
Llama3.1 on a raspberry pi
Thanks @jartine for Llamafile 😍
33
227
1,622
126,779
Llama 3.1 talks to your database for free with @OpenInterpreter
Why pay for a talk-to-your-database service?
Save money!
It's also fully offline and private, nobody else needs to see your data
25
95
669
44,276
No way... Now I can just tell my computer what I want and it does it?!
@OpenInterpreter
20
110
605
95,156
Here's a demo of @OpenInterpreter on my phone!
30
60
472
162,313
Control your computer with @jandotai and @OpenInterpreter
19
34
378
27,866
3 ways to use @OpenInterpreter for free, locally, with open source software in 3 minutes
6
29
183
16,805
Since you use Proton for email, it's super easy (and free) to use @SimpleLogin to generate new email addresses per service. Highly recommended
6
6
174
58,675
I'm all for building, hacking, and trying to make things easier for entrepreneurs in the age of AI but I have a genuine question -
How is this not a spambot?
6
149
2,775
Don't bother memorizing terminal command syntax
Just learn about the tools that exist and how to incorporate them into your workflow
Let AI do the boring stuff
Another free script to make you more efficient
9
12
114
12,405
Imagine an A.I. agent could act on anything you've seen or heard on your computer?
This would NEED to be open source and local!
That's why I integrated @OpenInterpreter with @screenpipe
It's not perfect but providing Open Interpreter with more context can unlock some powerful use-cases
5
17
121
11,525
Prohibition of harm against another person. With the definition of harm encompassing one's property.
The only thing we need a government for is to protect people from other people
2
85
6,761
For every A.I. startup that just got killed, a startup that leverages A.I. was just enabled.
The biggest leap since ChatGPT
4
2
80
12,737
Collaborate in the open for the good of the world, and still make bank. Literally no downside
1
80
14,774
Big news!
I’ve joined @OpenInterpreter full time
I'm so grateful that I get to spend my energy reengineering how humans interface with compute. Especially using open source software
Now, back to building the future
13
80
4,123
Here's a better demo of @OpenInterpreter on my phone! This time, I pass data to apps
9
11
83
34,865
OI will be in a billion devices
A small demo of how you can talk to Wingman AI in the future, thanks to openinterpreter.com/ #AI #WingmanAI @OpenInterpreter
2
8
72
12,101
Run thinking models locally!
The latest release from @ollama makes using thinking models super easy
2
23
73
8,551
One year ago today, I made a little demo of this cool new tool I found online. Just wanted to show off what it could do and then it went a little viral
Since then @OpenInterpreter has completely changed my life. I have made incredible friends, dove deep into the A.I. world, and am on a mission to accelerate open source tech
Thank you so much for being a part of this journey ❤️
Let's keep building an amazingly abundant future
10
5
73
9,803
Introducing the 01 Developer Preview.
Order or build your own today: openinterpreter.com/01
The 01 Light is a portable voice interface that controls your home computer. It can see your screen, use your apps, and learn new skills.
This is only the beginning for 01— the open-source foundation for this new era of AI devices.
3
1
66
1,807
Easiest way to run a local coding copilot!
Using @LMStudioAI and @continuedev together gives you a coding assistant without the need for an internet connection.
Don't lose your pair programmer when you lose connection
7
13
71
9,650
I'm so happy that I can finally share this!
Announcing Accessibility Grants
Technology should benefit everyone
Ohana and @OpenInterpreter are teaming up to help you build your ideas to make technology more accessible by providing up to $50,000 in AI credits
What will you build?
More details about grants and assistance in our blog post
7
7
63
12,031
Choose-your-own-algorithm seems like a no-brainer
Open protocol and data would be phenomenal
1
57
10,120
Llamafile in Colab!
We put together a notebook so you can run @OpenInterpreter in a free instance of Colab that uses llamafile to serve your inference
It's also a nice guide for how you can run OI fully locally
3
11
61
9,862
Do you use @LMStudioAI to explore Open Source LLMs? You should!
7
3
59
5,288
Try it with @OpenInterpreter!
interpreter --model anthropic/claude-3-5-sonnet-20241022
Computer use API
We've built an API that allows Claude to perceive and interact with computer interfaces.
You feed in a screenshot to Claude, and Claude returns the next action to take on the computer (e.g. move mouse, click, type text, etc).
3
6
59
4,996
Does your AI agent lose focus during a multi-step task?
Here's an approach that helps @OpenInterpreter stay on track PLUS it speeds up prompt engineering iteration
4
6
54
3,083
🚨 Announcing Anetic!
@FieroTy and I are building your interface to the future of AI.
Big things are coming, and it would mean a lot to me if you'd follow our new page. All future announcements will drop there, and you won't want to miss them.
In fact, we’ve already been tackling the #1 pain point we hear from AI enthusiasts: keeping up with all the latest AI breakthroughs is overwhelming and time-consuming.
So we’ve fixed that for you.
👉 Follow @anetic_ to learn how
6
13
46
13,395
Let's get tiny boxes running distributed training
What if you could use all the computing power in the world to train a shared, open source AI model?
Preliminary report: github.com/NousResearch/DisT…
Nous Research is proud to release a preliminary report on DisTrO (Distributed Training Over-the-Internet) a family of architecture-agnostic and network-agnostic distributed optimizers that reduces the inter-GPU communication requirements by 1000x to 10,000x without relying on amortized analysis, and matches AdamW+All-Reduce in convergence rates. This enables low-latency training of large neural networks on slow internet bandwidths with heterogeneous networking hardware.
DisTrO can increase the resilience and robustness of training LLMs by minimizing dependency on a single entity for computation. DisTrO is one step towards a more secure and equitable environment for all participants involved in building LLMs.
Without relying on a single company to manage and control the training process, researchers and institutions can have more freedom to collaborate and experiment with new techniques, algorithms, and models. This increased competition fosters innovation, drives progress, and ultimately benefits society as a whole.
This research is thanks to the hard work of @bloc97_ @theemozilla @apyh__ @UmerHAdil.
We invite researchers interested in exploring this area to join us in our quest.
1
34
10,699
Introducing the 01 Developer Preview.
Order or build your own today: openinterpreter.com/01
The 01 Light is a portable voice interface that controls your home computer. It can see your screen, use your apps, and learn new skills.
This is only the beginning for 01— the open-source foundation for this new era of AI devices.
2
3
45
1,800
Something huge coming in 2025!
Reach out if you love the idea of ambient agentic behaviour
After an incredible year at @OpenInterpreter, I'm moving on to start a new company with @MikeBirdTech.
I'm grateful for the experiences and learnings from @hellokillian and @OpenInterpreter, but now feels like the right time to start something new.
If you're obsessed with agentic systems and want to work with us, my DMs are open!
6
2
41
1,549
Accessibility matters
Ohana + @OpenInterpreter are committing up to $50k in AI credits to help the community improve accessibility!
Reach out or join the OI Discord if you have an idea that you want to turn into reality
2
11
41
6,584
Open Interpreter has hit 100 contributors! ❤️
Thank you so much for spending your time and energy on this project. It wouldn't be possible without you.
Let's get 100 more and blast this project into the stratosphere 🚀🚀
40
13,746
Making the 4 hour drive to Toronto!
Lots of time at the @cohere office
- Tonight, speaking at Local and Open Source AI Developer Meetup
- Tomorrow, Cohere Build Day
If you want to connect and talk @OpenInterpreter, AI agents, and open source, send a DM!
6
41
6,537
Open Interpreter September House Party nitter.app/i/broadcasts/1ynKODmvm…
4
2
38
4,731
10000 strong 💪🏻
Thanks for being the best community in open source AI ❤️
It's an honour 🫡
4
3
37
5,642
Lots of improvements with this release!
github.com/OpenInterpreter/o…
pip install open-interpreter --upgrade
6
38
4,791
Your open source, offline pair programmer!
Using @Ollama_ai to power @continuedev with the powerful model from @phindsearch
What a time to be a developer 🚀
6
5
37
3,163
Open Interpreter in a billion (fabulous) devices
What will your 01 be like?
i am opening preorders for a limited run of @OpenInterpreter powered Ai wearables discreetly placed in a Glossier You solid perfume metal compact.
to use, simply slide open to talk with Dua or to apply You, a signature scent from Glossier.
$129, preorder below ⬇️
3
36
2,659
Introducing the 01 Developer Preview.
Order or build your own today: openinterpreter.com/01
The 01 Light is a portable voice interface that controls your home computer. It can see your screen, use your apps, and learn new skills.
This is only the beginning for 01— the open-source foundation for this new era of AI devices.
3
2
36
1,872
People are free to take the software and make an app. They can also build any form factor of hardware they want
31
1,859
Introducing the 01 Developer Preview.
Order or build your own today: openinterpreter.com/01
The 01 Light is a portable voice interface that controls your home computer. It can see your screen, use your apps, and learn new skills.
This is only the beginning for 01— the open-source foundation for this new era of AI devices.
2
29
1,062
The most powerful tools we're released yet!
An intelligent @obsdmd plugin generator and script-o-matic to generate functioning scripts from a natural language description
I discuss the mission of @PrivSovCom and @FieroTy talks about Magic's 100m context window
Plus an amazing chat with @twodogseeds
I really like this episode
Tools that build Tools!
New episode teaches you about a script that makes scripts and an Obsidian plugin generator 🤯
Check out what @MikeBirdTech and @FieroTy made this week
Special appearance from @twodogseeds
We're finally verified so you can watch the full episode on X!
5
2
31
4,798
Super proud of this one!
I made a research assistant that takes a prompt and generates a comprehensive learning plan for you
(obviously open source, link below)
and you should see what Ty builds!!
Having so much fun on these weekly chats with @FieroTy 🫡
Babe wake up, new episode of Tool Use!
This episode is a 10x improvement, showing how to use web scraping for a prompt > synthetic data > fine-tuned model pipeline and an automated research assistant 🤯
All open sourced!
3
2
30
1,694
I don't plan on making a full app. Will share the full how-to guide next week so anyone can put it in their app. @hellokillian has inspired me to enable all the people!
28
959
Until you live up to the "open" in your name, people should continue to be skeptical.
Open source is what people should bet on
1
25
1,348
Is it old school to want to press a physical button to launch a voice controlled, fully capable AI assistant?
5
1
27
7,096
Open source, intelligent Obsidian plugin generator
From the latest episode of @tooluseai
Accelerate your @obsdmd vaults!
3
3
27
3,627
Open source Friends
Have you checked out what @kodjima33 is building?
Open source software and hardware will save the world
ALT open source hardware is going to change the world. What are you building?
5
1
27
4,841
Presentation in Ottawa about @OpenInterpreter and how to build the 01
piped.video/watch?v=sqwtk18p…
3
3
24
6,412
Screen cap: @screenstudio
Editing: @MadeWithCapsule
Transcription: @superwhisper
Terminal: @warpdotdev
Spotlight: @raycastapp
6
26
2,315
This is very exciting
Just updated, starting to play around
Ollama 0.2 is here! Concurrency is now enabled by default.
ollama.com/download
This unlocks 2 major features:
Parallel requests
Ollama can now serve multiple requests at the same time, using only a little bit of additional memory for each request. This enables use cases such as:
- Handling multiple chat sessions at the same time
- Hosting code completion LLMs for your team
- Processing different parts of a document simultaneously
- Running multiple agents at the same time
Run multiple models
Ollama now supports loading different models at the same time. This improves several use cases:
- Retrieval Augmented Generation (RAG): both the embedding and text completion models can be loaded into memory simultaneously.
- Agents: multiple versions of an agent can now run simultaneously
- Running large and small models side-by-side
Models are automatically loaded and unloaded based on requests and how much GPU memory is available.
1
1
26
983
If you want to teach a person something, you don't just tell them the answer
You explain the reasoning
Reasoning should be done openly and publicly so people can learn and not become dependent
2
2
25
403
<think>
You want things to be better than they currently are.
Wait, that means you have to enact change.
Wait, barriers are mostly mental and can be overcome with agency.
To change things, you must just do things.
</think>
You can just do things.
1
3
23
906
Try out @OpenInterpreter with Haiku 3.5
It's fast!
github.com/OpenInterpreter/o…
2
2
23
5,004
I used the @Raspberry_Pi 5 with the M.2 Hat with the Hailo AI module
3
24
3,120
A well orchestrated swarm of highly specialized fine-tuned open source models will be vastly more capable than a large generalized model
We have tools and infrastructure to build
1
3
22
4,419
Here's a playbook:
1. Familiarize yourself with Open Interpreter. Play with it, make small Python scripts that use it, join our Discord to learn about it more deeply
2. Break down the product idea into small tasks. Then do the same with each task. And repeat one more time. You want to have a very through to-do list
3. Get building. One task at a time with periodic review of the to-do list. Your initial list won't be the final list but the only way to figure out the final list is by building. Doing this in public gets you assistance and generates buzz
4. Get people to try it and take feedback. One note about feedback, people are almost always right about what is bad about products but are rarely right about what would make it better. Nobody will know your product better than you but nobody knows how it's bad better than everybody else
5. Keep grinding and enjoy the process.
Let us know if you have any questions and I'm excited to see what you build!
2
3
24
2,270
The goal is for open source to win
- Connect people who will amplify each other's efforts
- Share knowledge
- Make contributions
Keep being a team player and this team will win
2
3
23
689
Inference: @ollama
Terminal: alacritty
Database: postgreSQL
Screen cap: @screenstudio
1
1
22
1,974
Check out this 01 case model walkthrough from James! He made some great changes to make the 01 his own
piped.video/watch?v=BjoO0Kt-…
Love seeing this project evolve
Open source will win
19
1,828
The love and support that I've been shown this past week has been incredible
Thank you ❤️
Will definitely make you proud 🫡
3
19
731