"I get taught life lessons all the time. Probably the most important one is that people will tell you [that] you can’t do something, and you have to ignore them because you can."
"Are you working to become a good ancestor -- to make something truly interesting -- or are you just working to make money?
Because you can make money by burning villages down and rebuilding them; that’s a good way to make money, but it’s not a good way to be a good ancestor."
"If you don’t contribute, don’t pay attention and you don’t give back then it can go away in the blink of an eye...
And so I’m rededicated to this notion of no: if you want a democracy it doesn’t mean you live in a democracy.
It means you participate in it or you lose it."
"Design is not art.
I was not an artist; because an artist has a story to say, to tell, and I didn't. I was a craftsman. I loved the tools and the trappings of art.
I realized that there was a different role for me: and that's the thing about design. Design is problem solving."
"Some people just really like to create things themselves.
I think that's why I've never built a big company -- because at some point, you have to stop making things, and you have to start working through others; and I'm much more interested in the things, than the others."
• • •
Missing some Tweet in this thread? You can try to
force a refresh
👋 Inspired by recent conversations with friends, and based on a long history of automating away every job I've ever had (from data processing to PM work):
Am sharing a few ways that I'm using Gemini 1.5 and 2M+ tokens of context in @GoogleAIStudio to automate the boring parts of DevRel and UXR!
Reminder that you can stuff quite a bit into 2M+ tokens (hours of video, years of emails, full codebases, etc.) and that, over time, we expect 2M tokens ➡️ infinity, cost ➡️ $0, latency ➡️ near instant.
(1) Uploading a dated codebase (in this case, Flax 0.7.5), and a newer version of the codebase (Flax 0.8.5), then analyzing changes.
You can generate documentation changes based on the differences in code; blog posts or release notes describing the code changes; and - a favorite - update old tutorials based on the new versions of the APIs.
(2) Analyzing product feedback at scale by scraping @GitHub and @Gitlab issues, conversations in @Discord and @Discourse forums, social media chatter, etc.
In this example, I scraped a whole bunch of user feedback about the OSS vector database, Chroma, and compared it to feedback on a competitor's tool (Qdrant).
✨🤔 Wondering how far a person can get with "make this code faster", "make this code more readable and reusable", "refactor this code to be more concise" in the prompt.
👇🏻Am also impressed Bard deduced that I was attempting to implement a multiplication table!
✨👩💻 Jazzed to imagine a future where we all have friendly, competent technical assistants that cheerfully answer n00b questions about chemistry, physics, math, and programming.
📝 Citing sources would be a strong next step, just as we cite potentially recited code in snippets!
✨Bard even recommends unordered maps instead of ordered maps in C++!
📊 Is anyone else *super* dissatisfied with the tech industry's preferred/tracked open-source metrics?
@github stars; pip install or download counts; @-mentions or tags on social media: all of these stats can, and will, be gamed. We can do much better!
👇🏻Here are some ideas:
@github (1) Projects listing a particular repo as a dependency.
This can be easily tracked via GitHub's dependency graph, or by scraping which Dockerfiles, conda environment YAMLs, etc. reference a library or framework.
(2) "Bus factor" of a particular open-source project.
Bus factors measure how resilient a project is to sudden engineering turnover - and is a solid method of understanding the health of an open-source project.
🤖 Reinforcement learning in production is a very nascent space, but a fast-growing and multi-faceted one (everything from game dev to operations research)!
👇To showcase this, am compiling a list of projects that are using @raydistributed and RLlib to enable their experiments:
Everything from multi-agent reinforcement learning; to game balancing and boss optimization; and (even sometimes outside of the realm of RL, but still powered by Ray): in-app game recommendations.
⚡️ This scenario is very near to my heart. Did you know that you can optimize electricity use in a plant or a home; model thermal grids; and manage energy resources efficiently using RL models?
The longer I work on open-source ML tools, the more convinced I become in decoupling libraries.
Crafting simple, delightful, and composable user-facing APIs is *endlessly difficult*; you shouldn't also have to have a PhD in distributed systems in order to make those APIs scale.
Library authors should be able to focus on building concise, extensible features for their users–that help domain experts go from having an idea, to realizing it, as quickly as possible.
Asking those authors to worry about hardware, or data / model parallelism, is unreasonable.
And having to communicate to a user that (as an example) an image preprocessing feature that worked in one framework, won't work in another–
that they have to hunt down an identical transformation, in the context of the new framework–