Joe Heitzeberg
@jheitzeb
Followers
6K
Following
7K
Media
1K
Statuses
9K
AI Tinkerers global, co-founder of Crowd Cow, 2019 Allen School Alumni Impact Award, MIT, weekly 1:1 with my AI
Joined June 2007
reminiscing on an AI Tinkerers sidebar chat a couple of years ago. When it's early, nobody knows! these debates (followed by tinkerering and buiding) are how to find out if ideas will work
4
1
8
Asked “who has switched from ChatGPT to Claude” at AI Tinkerers Seattle tonight, approx 60-70% of hands went up.
31
30
413
Wow! Ultravox is an *open source* speech to speech model — understands non-textual speech elements — paralinguistic information. @juberti just showed how it can pick up on tone, pauses, and more! @AITinkerers Seattle @FixieAI
14
52
365
Correctly answering questions from a 14,000 word scientific paper… pointing my chatbot to a doc store with embeddings index
22
23
319
I made something that can take two input: YouTube video url + arbitrary text — and do this
36
27
267
Who else uses GPT-3 several times per day as a Google replacement?, e.g.
9
8
225
I created an AI-generated podcast to help me learn and keep up with topics I care about, using AI to find, read and summarize articles it thinks I would I care about, and to generate a podcast in my own voice using voice cloning from @elevenlabsio 👇.
13
33
227
Read all the hacker news comments on this page and summarize it for me
10
15
173
AI browser control in Ruby! Inspired by @sharifshameem a while back, this uses GPT-3 prompt chains to control a browser. @natfriedman said I should release the code, so here it is: - hope to allow Ruby devs to explore something that feels magical :)
6
23
154
Well on my way to delegating lots of reading and understanding to my AI agent. Next up: ingest a ton of papers, blog posts, news articles, prompt with the things I care about, have it summarize, synthesis and bring the most important things to a regular "1:1" with my AI.
9
8
125
"The simplest thing I can build that can build itself" . — that’s the guiding principle of Yohei Nakajima @yoheinakajima, the mind behind BabyAGI, as he challenges the boundaries of what AI can achieve independently. In my latest episode of @AITinkerers "One-Shot", we dive into
8
27
120
Mind = blown! @rahulchhabra07's Neuralace built a model from.EEG to… read the mind. Absolutely mind-boggling! #AITinkerers #SanFrancisco
8
19
98
ChatGPT is (a) remove login, (b) get a better name away from being the first serious Google search competitor in. ever.
10
6
76
Keeping up with popular open source AI repos is hard, but we've got you covered! Blueprint AI summaries of Auto-GPT, @LangChainAI, @babyAGI_, @gpt_index and more 👉 cc @hwchase17 @yoheinakajima @jerryjliu0 @SigGravitas 🔥.
2
8
58
Correctly identifying guitar sounds with Google Gemini 2 voice-to-voice
Great to have this repo! thanks @kwindla! Gemini 2 Voice-to-Voice = the fastest voice interface? Pipecat means fast transport. It's clearly hearing non-verbal audio but struggles to describe it, and can't sing or laugh. Great for practical apps. This repo rocks!.
4
8
58
AI Tinkerers San Francisco - first of 8 demos (of 49 proposals, > 300 wait list) — active builders! With @natfriedman @alexgraveley @mickeyxfriedman @aigrant_ and 225 tinkerers!
6
2
56
The next AI Tinkerers Seattle Meetup is May 11th sponsored by Ai2 Incubator: "On the Road to AGI with Agent Systems" featuring @yoheinakajima and community demos - RSPV here:
9
12
54
Using a fine-tuned Moondream on street view image to do less gen for house painting services!
2
4
51
100% open source AI Artifacts clone by @mishushakov build with @e2b_dev and running @MistralAI at @aitinkerers Paris
3
8
52
Diving into agents that build their own tools and a vision for sharing and competing those tools, from BabyAGI to BabyAGI2 and beyond with @yoheinakajima, @AITinkerers One-Shot episode coming soon
5
4
51
Sup? This:. The first AI Tinkerers hackathon in SF and it's Nov 2nd + 3rd --- we have founders of @e2b_dev and @humanlayer_dev, Gemini engineers from @Google, the engineer who built agents for @Replit, the CEO of @browserbasehq , the founder of @trydaily , founders of.
5
12
50
The @AITinkerers Humans-in-the-Loop.Hackathon is officially underway! • Excited to see the innovation unfold over the next 48 hours. Big thanks to @Anthropic, @CopilotKit, and all our sponsors! @digitalocean @e2b_dev @neondatabase @humanlayer_dev @MadronaVentures
1
16
49
I created a work flow that takes any short video clip and automatically creates a viral video version of it in one go. Here's how it works. 1) upload the short original clip [me explaining my espresso machine]. 2) extract the final frame and ask moondream to analyze it and come
6
5
47
The @AITinkerers Summer Soocial 100% active builders and AI is on - Seattle with @GroqInc @MadronaVentures @Microsoft and friends!
3
3
42
I'm getting lot more useful real work done with Gemini 1.5 and 1m tokens by jamming in massive amounts of context than I've ever gotten done with RAG and GPT-4.
7
3
37
1
1
39
“Stable diffusion was the first thing that gave me hope that there would even be an open source large AI thing at all.” — Vik from @MoondreamAI 🔥. In this One-Shot moment, Vik shares how open-source vision models pulled him back into ML—and how YouTube (hi Karpathy & Jonathan
4
8
38
A lot of people have been playing with this (embeddings + dynamic prompts) to implement document Q&A for a while. GPTIndex @jerryjliu0 and LangChain @hwchase17 are two libraries that can help with this.
1
0
35
Hats off to @ronaldmannak for making it easy to run LLM models locally on Mac with Pico MLX Server (apples native architecture for running models)
3
6
37
Answering questions from a 106,000 word document in French. The question and all sentences in the doc are embedded, several results (and surrounding text) are brought back and fed to a prompt that is asked to determine the answer (if possible) from the snippets.
4
0
35
One of the most narrowly useful apps I’ve built and use all the time. Paste any url and it find the PDF, downloads it, reads the first part and infers the title, renames the file and sends to my kindle.
4
0
34
@zackkanter lol, even my 12yo (who just saw this tweet and headline in my feed) instantly laughed and said "nobody will forget chatgpt. who wrote that article, sheesh".
0
0
34
Try it here… (the first question builds the index, so be prepared to wait or to retry if there is an error).
2
2
32
🔥🔥 Momentum! founders and experts from @e2b_dev (secure sandboxes for agents), @PaymanAI (AI-to-human payments), @browserbase (cloud browser infrastructure), and @trydaily (real-time voice/video for AI) are joining the @AITinkerers Hackathon on Nov 2-3 with @googlecloud at
3
8
32
I’m truly grateful to have experienced 1/3 of my life without computers / internet and now I get to experience another 1/3 with AI. Truly an amazing time to be alive and in the thick of it!. I mean, roughly speaking :).
6
4
30
This Is currently the most excited any AI has ever been 😂 meta raybans giveaway at @AITinkerers 😂 @weights_biases @altryne
4
4
31
Gemini 1.5 Pro is amazing. I have a (private) medical application that I build with a RAG stack (medical records, PDFs, clinical notes) -- 450,000 token prompts that Gemini single-shots with better results, more quickly too. Amazed with results so far.
3
1
30
Indexing was surprisingly fast. I indexed per sentence. For retrieval I surface the best 3 matches and the put each, together with the pre and post sentences (for more content) into a prompt that determines how to answer.
3
0
28
Thank you @flexport @typesfast Ryan! for hosting AI Tinkerers - SF tonight! s/o @composiohq @SignalFire with @dexhorthy the organizer
3
1
29
4
3
28
Btw, I’m not using any vector db. just embed the question, pull all the indexed vectors and compute cosine similarity, sort and pull the top 3. Hundreds of embeddings per doc. By far the slowest step is then asking gpt3 to compute an answer given the top results and the question.
5
0
26
Announcing: Humans-in-the-loop Agents Hackathon in San Francisco - November 2nd and 3rd with @googlecloud @ @weights_biases - come build agentic apps and win prizes @AITinkerers first hackathon in SF!
4
5
28
lol RAG is not dead. It already evolved to agentic RAG doing iterative reporting and actioning along the way. Besides, 10M only fits 8 years of iMessage history and probably a pretty small fraction of a typical inbox.
7
1
27
Paul from @browserbasehq kicking off @AITinkerers SF! Incredible community demos lined up, and folks from around the world
2
2
27
My demo at AI Tinkerers tonight was an open source workflow to create viral videos in one shot using Minimax and others on @replicate
1
4
28
I spent a month having coffees with a few of the AI Tinkerers members. Feedback was loud and clear: what makes AI Tinkerers different and special is two things.1) the people: literally everybody is building something.2) the focus on technical details, not pitches."The best meetup.
0
3
25
Me: "Email Jonathan" .Siri: "what?" .Me: "Email Jonathan" .Siri: "Now playing Coldplay on spotify".
4
0
25
At @AITinkerers Paris with @htahir111 from Munich who is AI Tinkerers Seattle member’s brother @deepwhitman — small world!!
1
1
25
To hear the result, go here: To create a podcast, I just enter topics I care about like "AI, GPT-3, Seattle" and 1 minute of my voice as an MP3. The rest is automated and runs 3x / week.
2
0
24
Quality of demos at tomorrow's second AI Tinkerers Bangalore is 🤯 -- don't sleep on India for GenAI 🇮🇳 🚀.
5
3
24
Tomorrow night - San Francisco, Seattle and Austin have AI Tinkerers meetups happening at the same time. Here's a preview! 👇. SF: The talks include virtual avatar generation, a code-understanding wiki, web data extraction, AI-powered scheduling, open-source Anthropic Artifacts,.
1
4
22
🚀 Join the AI Tinkerers Seattle Summer Hackathon on June 10th! Sponsored by @awscloud, @CohereAI, @MadronaVentures & supported by @OpenAI, @pinecone, @weaviate_io, @statsigio, @getblueprintai. Amazing prizes, networking, food & drinks await! 🏆🍻 🤖. Sign up & spread the word:.
0
7
22
If you’re making agents and not using @browserbasehq are you even making agents? AI Tinkerers SF
1
4
24
Please, tell me about your marketing automation startup idea and how you will differentiate from the competition [landscape graphic]
1
2
20
Going around sharing the insights from tonight’s @AITinkerers hackathon pre-function dinner conversations @charles_irl @jamescham’s moderating and new friends like @kwindla @composiohq @restackio @dylan522p @CopilotKit
4
5
22
Updated AI Summarizer to use the new ChatGPT API -- holy wow, much faster, and results are better too!
4
3
22
@emollick Yes - but most roles in companies are fairly narrow, repetitive and don’t offer a ton of autonomy - the level of reasoning, decision-making and autonomy needed by AI to automate many roles may not be too far off — also the fixed cost of setting up an automation is worth it —.
1
0
21
195 signed up to AI Tinkerers Seattle Summer Social since 9am this morning — a good start to what will be an epic gathering of 100% active builders, with @GroqInc presenting their amazing latest and community sponsors coming onboard to tinker during the networking s/o.
3
3
22
hey @Google for goodness sakes, Microsoft is building LLM features across their whole suite, and OpenAi is eating your lunch. If you're so worried about the brand impact of chatbots that swear, then start with very useful narrow features like autocomplete gmail w/ calendar:
5
0
21
I love this direction for @LangChainAI - would love to see more vertical app specific tools added!.
Introducing the 📧Gmail Agent Toolkit📧. 📨Enable agents to search Gmail, retrieve, draft, and even send messages!. Docs:
0
3
21
🔥 wowzers this is freaking fast!!🔥
Today we’re thrilled to introduce Ultravox v0.4.1, a family of open speech models trained specifically for enabling real-time conversation with LLMs. No ASR step required. Ultravox's speech understanding is the best in open-source and is quickly approaching the quality of GPT-4o
0
1
20
Realtime vision with moondream - confirmation that petting of dog earns points in the latent space @vikhyatk @dksf @ConwayAnderson AI Tinkerers SF
4
1
21
Had a product idea 7 days ago and now I spend 7 hours a day thinking about it to the determent of all other work and yet also can’t find anyone building it… must I built it??.
11
0
21
@sharifshameem @sergeykarayev @goodside @amasad I build a prototype of this with headless chrome in the cloud, a simple DSL and some logic with prompt chaining… feels like magic
0
0
21
Everybody’s cooking with @Google Gemini @AITinkerers @shresbm @IlNardo92 AI Tinkerer HITL Agents Hackathon coooooook!!!!
1
2
21
We've gathered an exceptional set of speakers for this such as @alexgraveley @hwchase17 @spolu and @irinarish - now @OpenAI is speaking too, and there are prizes like $5,000 in OpenAI credits! check it out!.
🚨 We have an exciting announcement to share: @OpenAI will be speaking at Launchable: Foundation Models! Participants will hear from Head of GTM @iamthezack, and get special access to OpenAI credits and AI experts. Apply now: @madronaventures #ChatGPT3
0
4
20
I added in the AI Connections — a dedicated agent reviews each job and scours the network doing outreach only when the AI Tinkerers member is a perceived perfect fit based on specific technical abilities and fit - good traction so far
4
2
20
Heading to Chicago in a couple of weeks for AI Tinkerers - Chicago - heard it’s going to be a special one!
3
3
19
Just added a browsable directory to all @AITinkerers cities to facilitate connections between attendees
3
1
18
AI Tinkerers is coming to Paris, Prague and Dublin -- all in the same week. I'll be there! Register now, link in comments @AITinkerers.
1
2
20
AI Tinkerers - Palo Alto inaugural meetup with Hippocratic AI founder @munjalshah - one of the most important AI efforts
1
1
20
Huge congrats to @pk_iv on @browserbasehq's $21M series A! I sat down with Paul for an episode of "One-Shot" to dive into Stagehand, their open source browser control library released today! 🅱️ ~enjoy!
0
2
20
Just released a new "One-Shot" episode featuring @KaranVaidya6, co-founder of @composiohq!. We discuss the power of AI agents and why API integration is crucial for secure, scalable automation instead of relying on risky browser-based controls. Check it out! 👇.🎥
2
5
20
To kick the year off right, I woke up, read (part of) a book, took a nap, around noon sat down to a blank slate and built in its entirety, exercised 1.5 hours and had dinner. Off to California in the morning.
4
2
19
Check out @sdand's Numpy-powered vector database at “ye” combinator — fastest vector DB. #AITinkerersPaloAlto
3
2
19
This is pretty much the magic of @AITinkerers right here - agent systems founder by way of Chicago and SF asking Paris agent founder how he dealt with ambient agent system concurrency issues and gets a technical and useful answer straight up @mehdidjabri 🔥.
1
1
19
“You put something out there and people educate you on how it can be used.” —@MoondreamAI 🧠. Vik shares how a Texas rancher used Moondream + drones to track escaped cows—and how companies are captioning millions of images locally for model training. Watch now:🎥
0
1
19
Joe, Joe and Joe! @JoeGoJoeGo @joechoigreene there were actually at least 4 Joes there tonight @joevandyk
0
2
18
Running llama 3.2 (3B) locally on iPhone Pro with the PocketPal app…completely unthinkable a short while ago
2
2
20
@natfriedman I prototyped something similar! Another approach is machine vision to identify the UI elements and text, to void the dom/verbosity issues. I also did the html simplification, but loses much useful info such as class and other queues the LLM could use.
1
0
19