Opens profile photo
Follow
Joe Heitzeberg
@jheitzeb
Entrepreneur, computer scientist, MIT Sloan MBA and co-founder + CEO of Crowd Cow. 2019 Allen School Alumni Impact Award, weekly 1:1 with my AI agent
Seattlecurrentlyobsessed.comJoined June 2007

Joe Heitzeberg’s Tweets

Schools banning ChatGPT? Instead of banning technology, here’s an example of how schools can adjust to the reality of generative AI. Consider a high school history class, and a typical assignment: to read a history book, and write an essay… (thread)
34
465
Show this thread
If the costs were not 😶 I'd let everyone create one, but the tokens and tts costs are non-trivial given how much info I am combing through to create this. Worth it for me though :)
4
2
Show this thread
Then I pull down the full text of those articles and summarize them recursively (note: they're often far longer that would fit in a GPT3 prompt), picking out key points and learnings
2
1
Show this thread
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 👇
11
222
Show this thread
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.
Image
5
37
Show this thread
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.
Image
9
133
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
28
Show this thread
For long docs, this makes answering detailed questions work very well. But if I want to answer general questions pertaining to the entire doc (in the doc, how many times is XYZ mentioned) I’d need to look at more results, or also embed summaries.
1
17
Show this thread
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.
4
31
Show this thread
Instead of banning AI in education, schools should teach about AI, how it works, and how to use it as an amplifier of human potential. If we ban students from using computers, the Internet, or AI, what future career are we preparing them for?
29
389
Show this thread
We've gathered an exceptional set of speakers for this such as and - now is speaking too, and there are prizes like $5,000 in OpenAI credits! check it out!
Quote Tweet
🚨 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: bit.ly/3FcW9dN @madronaventures #ChatGPT3
Image
20
hey 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:
Image
4
21
When I'm solving problems as a human, the more I consciously listen to my inside voice and take notes about how I work through problems, the better I become at LLM prompt engineering.
2
11
I'm reading every week in 2023. Advice threads, GPT-3 demos, war assessments, shitposts, or anything people like a lot. I'll keep adjusting the list. Start on Monday, done by Sunday. Might make lowkey videos of takeaways. If you want to read along, the current list:
Image
5
259