Dancer. Cofounder @typesafeai - all things model capability. Prev: @AIatMeta
Mission, San Francisco
Joined August 2011
- Tweets2.3K
- Following2.1K
- Followers9.7K
- Likes5.2K
Pinned Tweet
Hello Software Engineers, I lead TypeSafe AI Model Capabilities team. We are hiring. Here are a few qualities we are looking for:
• strong plus for people who resonate with the title "Experience Engineers": mobile engineers, full stack engineers, or even (virtual) reality engineers.
• strong coder (you need to be able to manually write code)
• product oriented thinker
I promise are fun! I trust that you will figure out how to apply.
Happy JevDay 10.14
• v1.14 with better multilingual support
• Image support in state, instruction, AND criteria
• It returns 1 bounding box!
• longer context up to 262k
• OPT-IN flag for us to learn from you!
Happy Hacking
Happy JevDay!
But before announcements we have to talk philosophy…
The needs of early adopters / tinkerers / dreamers are quite different from what is needed for extremely stable infrastructure.
For the Jev release, we erred on the latter, but this release is for the former - with a gate so people can opt-in to the danger zone!
Tune in to our town hall in Discord at 1:14pm
discord.gg/typesafe
what's this?
Jev is evolving!
tune in to our town hall with @CompleteSkeptic and @allietheicon at 1:14 PM to learn more
You own what you create!
was at a community event and was asked a somewhat sad question: should developers be afraid of the labs eating the new possibilities jev opens up?
with AI (system one) + software (system two):
1. there's no going back to a world without system one models
2. you control the software!
3. there's no way AI could beat the efficiency of this combo
4. you don't even send the model providers your whole program logic
I think the days of "wrapper" apps will be seen as some of the darkest in software because people essentially did all of their programming in a system message and sent it over to a third party who was incentivized to bake that behavior into their 1st party product...
keeping your logic in code and only asking AI for the minimum reverses that and keeps all of your logic yours. if you decompose enough, model providers can't even figure out what you're working on
Series A, LFG!
launched a model
accidentally served trillions of tokens a day
29.4% of the fortune 500 showed up
raised a really big series A from @a16z
it's been 3 weeks
we would like to sleep now
If you don't super understand how to use Jev - you can check out our cookbooks. We spent a long time writing them. docs.typesafe.ai/cookbooks
If you are interested in ZDR or want higher rate limits, we are offering that through [email protected]
Happy Hacking
Agents on Jev - straight into production
TypeSafe AI 🤝 MotherDuck
One week into shipping direct Jev support in MotherDuck SQL and we have (1) maaaaaany billions of tokens generated and (2) way more customers running real production analytics pipelines using prompt_jev() than I expected!
What are real customers using prompt_jev() for?
- The top use cases are tagging product and event data of varying shapes, and labeling and scoring support tickets, free-form text, and of course sales calls.
- Agent productivity is more common than I expected. Lots of customers have tons of agent traces and want to tag them by what they're about. Everyone's trying to get to the bottom of dev productivity & new product analytics w/ their own agents.
- Customers of all sizes and stages use it. Big, established accounts, small nearly-dormant ones who don't want to talk to sales springing to life, everything in-between, and tons of new signups who upgraded to our Business plan and immediately started cranking.
I'm going live tomorrow (Tues Sept. 29th) at 9:30am PT with @matsonj to dig deeper. Link in reply
People on discord are so cute - they were concerned about my personal info leaking from the onboarding lesson. I also randomly jump in voice chat sometimes and contribute - @discord is where all the actions are. Join us there.
(And pls send me les memes)
