Early stage investor at Innovation Endeavors, focused on computing infrastructure, data/AI, and tools for builders.
San Francisco, CA
Joined September 2020
- Tweets358
- Following5.4K
- Followers1.2K
- Likes1.9K
Davis Treybig retweeted
For a long time I’ve talked about Agent Experience. Many think AX is the new DX, but that misses the bigger picture.
We’re moving to an agent-first world, where agents are the primary users of all software.
AX isn’t just the new DX. It’s the new UX.
innovationendeavors.com/insi…
Davis Treybig retweeted
a little update from Extend
turns out the world has a lot of PDFs
(before the y-axis police arrive: we're in 8-figure ARR territory)
Davis Treybig retweeted
Meet the Capsule Video Agent. Available now.
Yes, this launch video was made with it (in 3 shots).
No, this isn’t just HTML pretending to be video. We built a truly native video rendering engine from the ground up that outperforms HTML in aesthetics, speed, capabilities, and portability.
For the first time, making a professional, on-brand video is easier than writing a doc, building a deck, or designing a graphic.
I believe the impact on how we communicate will be profound.
To put it in context: roughly a billion people make slide presentations every month.
Until today, only a tiny fraction of those people knew how to make a video.
Now every sales follow-up gets a personalized video.
Every prospect gets a custom demo.
Every proposal becomes a visual pitch.
Every feature gets a video when it ships.
Every blog post becomes a YouTube video.
Every customer question gets a video explainer.
Every new employee gets personalized onboarding.
Every training document becomes a video lesson.
Every company announcement becomes a video.
Every QBR comes to life.
From this day forward, every idea worth sharing becomes a video.
It's a big promise, I know.
But give it a try.
I think you'll love it.
Davis Treybig retweeted
What happened with agentic coding last year is happening to video editing right now.
No one will use a traditional NLE like Premier or even Capcut a year from now.
Huge opportunity to reimagine the surface area and workflows for both creative professionals and the millions (billions?) of people who will finally be able to make professional quality video on their own.
A few thoughts on agentic video editing I’ve had over the last few days:
1. 99% of ads should be edited agentically now. If your editors are still editing in the timeline, they’re falling behind. I got a message from my senior VE saying that he’s putting out 5x edits/day now at the same quality he was before.
2. The conversation with video editors will be uncomfortable because the timeline is all they’ve ever known. Some will be hesitant and some will be fearful for their job. I think the best way to address this is head on. The tech is inevitable, but having taste is still going to be incredibly valuable to train and give feedback to agents. Those who have this will become more valuable, those who don’t will struggle.
3. Opus 5.5 can natively edit all of your ads without needing any MCPs or plugins. I’ve tried the alternatives and couldn’t find anything that was significantly better than doing everything inside of Claude.
4. Have your senior editor(s) build skills for each style of edit (e.g. mashups, UGC, AI animations…). You can build these by feeding 4-5 inspo edits to Claude and tell it to make a skill out of it. I prefer building lots of mini-skills to one big video editing skill, so that the context stays as relevant as possible.
5. Once you have a base skill (e.g. ugly ads edits), the best way to train it is through feedback loops. Have the skill re-edit the last 25 ads each senior edited then put them in an artifact, and have the senior VE give feedback on each. Keep doing this until the skill is refined and it understands what a good edit is.
6. If you do this, there’s no reason why you shouldn’t be able to make edits that need little/no feedback. Most of ours are approved after v1 or v2, sometimes v3.
7. Only A-player editors should be training the skills.
8. Re adoption in the team, the AI champion builds the v1. Then get out of the way. I’m not the best editor on my team, so my job is to build the harness that lets the best editors train the skills, not build the skills myself.
9. If you want to supercharge this process, have your top editors record their screen as they edit. Turn on Gmeet or loom and get them to narrate the reasoning behind their process. E.g. why they chose this cut, pacing or b-roll clip…this is all context that gets fed into your skills.
10. It takes time to build, train and refine these skills. You have to go through the dip to get out the other side. Budget time in for senior VEs to work on this - otherwise you’ll end up with skills that put out subpar edits and don’t save you much/any time.
11. Operate under the assumption that this is the worst video editing is ever going to be. Opus 5.5 can do most of our editing today…what will Opus 6/Fable 6 be able to do? The editing skills you build are only going to become more valuable as the tech evolves.
I still don’t think people realise how big video editing being unblocked is. Things that I thought were 3-6 months away are now on the table.
Will continue to share more as I learn. CRAZY, crazy times.
Davis Treybig retweeted
Design is NOT dead.
Anyone can generate an interface now. Designing a great experience still takes empathy, systems thinking, and taste.
Introducing Soffi Sessions 🎙️
In our first episode, our Founder & CEO @crisdobbins shares why design isn't disappearing. The canvas is just getting bigger.
Watch 👇
Davis Treybig retweeted
1/4 Today, we’re releasing Mercury 3. Mercury 3 is by far our most powerful model and the most intelligent diffusion LLM ever released. It outperforms GPT-6 Luna and Haiku 5.5 on quality while running at 1000+ tok/sec.
🤖 Made with AI
Davis Treybig retweeted
AI-generated fake documents grew 500% last year, and humans can no longer tell the difference
That's why we’re launching Extend Verify, the first document fraud detection API built for agents
Verify catches the fastest growing types of document fraud, including:
- fully AI-generated documents
- localized AI edits (down to a single modified character)
- patched text, overlays, and in-painting from Photoshop, PDF editors, and coding agents
- digital tampering after a document was signed
- and many more
In our internal evals, Verify detected over 96% of AI-generated and tampered documents.
Verify is available today: extend.ai/verify/request-acc…
Davis Treybig retweeted
Bit more than a decade in training, no pain no train. 💪
@BEBischof nerd sniped me into this talk with offer to have a Cat to give the keynote.
Still Training: A Decade of Learning the Hard Way @ Reverie by LanceDB on Nov 5 in SF 🌉
You get to hear me about distributed training from the trenches:
Hardware codesign, async and sync training -- when we almost decided to put async into chips
From big batch gradient descent with parallel boosting from stochastic optimization with duals to online learning where we had regrets, to higher order methods that scale that was almost second order that nobody believed it.
This is genuinely one of the most unbelievably magic products I have used in ages. It lets you make almost arbitrary videos about anything - explainers, tutorials, launch videos, silly vignettes. The consistency and quality is hard for me to explain - see for example this "Mac vs PC" style video about open vs closed weights
Meet the Capsule Video Agent. Available now.
Yes, this launch video was made with it (in 3 shots).
No, this isn’t just HTML pretending to be video. We built a truly native video rendering engine from the ground up that outperforms HTML in aesthetics, speed, capabilities, and portability.
For the first time, making a professional, on-brand video is easier than writing a doc, building a deck, or designing a graphic.
I believe the impact on how we communicate will be profound.
To put it in context: roughly a billion people make slide presentations every month.
Until today, only a tiny fraction of those people knew how to make a video.
Now every sales follow-up gets a personalized video.
Every prospect gets a custom demo.
Every proposal becomes a visual pitch.
Every feature gets a video when it ships.
Every blog post becomes a YouTube video.
Every customer question gets a video explainer.
Every new employee gets personalized onboarding.
Every training document becomes a video lesson.
Every company announcement becomes a video.
Every QBR comes to life.
From this day forward, every idea worth sharing becomes a video.
It's a big promise, I know.
But give it a try.
I think you'll love it.
agent.capsule.video/ please try you will find it addicting
Davis Treybig retweeted
introducing operator-1, an autonomous document extraction agent
operator-1 saturates multiple 3rd party benchmarks, and delivers the most accurate performance on complex tasks that other solutions can't handle:
> long tables with thousands of rows
> lengthy files with context scattered across hundreds of pages
> dense diagrams and images
> much more!
operator-1 is a specialized agent we built that:
- runs within a custom workspace
- knows how to use dedicated tools (parse, search, zoom, code execution, validation) to handle any document
- learns via persistent memory
Davis Treybig retweeted
We ran the new models through our coding agent token usage benchmark. @OpenAI's GPT-6.1 Sol vs @AnthropicAI's Claude Sonnet 5.5 and Claude Opus 5.5.
For every one of them, the biggest line item was context. Claude still does more formal planning than GPT.
GPT-6.1 Sol put 58% of its cost into context and only 3% into planning. It finished all 5 tasks for $0.45.
Sonnet 5.5 and Opus 5.5 spent 40% and 38% on context and about a quarter on planning.
Opus cost $1.58 (~3.5x Sol). Sonnet cost $0.72 (~1.6x Sol).
Davis Treybig retweeted
OpenRouter for generative media is here 🦾
Comfy Router is live
One API for frontier image, video, 3D, and audio models.
Same model string. Same arguments. No new SDK, no new key, no redeploy.
What Comfy Router gives you:
→ Explicit routing. You name the provider, we call that provider. It's down? The request fails there. No silent fallback.
→ Every job returns the provider that ran it. Log it, bill it, debug it.
→ Async. submit() returns a request ID immediately. The queue retries 429s and transient errors until a slot opens. subscribe() submits and polls to completion.
→ Batch-friendly. Queue a few hundred jobs, hold the IDs, pull results as they land. Nothing blocking on a 5-min video render.
→ 24h retention on inputs and outputs, then deleted.
→ Comfy credits. No sub, no Router fee.
Providers at launch: Comfy. Runware, Wavespeed, Fal, Higgsfield. Multi-provider where the model supports it.
Get Your API Key with the link below. ⬇️
Davis Treybig retweeted
rare is the team doing true architectural AI research these days. I talk to the extraordinarily broad and productive researcher Stefano Ermon about his company, latency, and why it’s still worth going after good ideas in the age of scaling
Today's LLMs can't generate token 10 until token 9 exists.
@_inception_ai CEO @StefanoErmon thinks he's found the winner: diffusion LLMs that generate tokens in parallel.
New episode of No Priors, out now.
Davis Treybig retweeted
today, we're launching our new Parse Router
extend now runs a classifier on every page, and decides whether to run our performance or light engine for parsing (on a page-by-page basis)
tldr:
- most documents have a mix of complex pages intermingled with simpler ones. The router classifies and routes each page so you get the best blended performance x cost for any given document
- you pay per page for whatever was used: 0.5 credits for our Light engine, 2 credits for Performance
- our classifier works by scoring each page on layout, text quality, tables, form fields, and checkboxes
- on RealDoc-Bench, enabling the parse router scored 89% (close to Performance's 90.9%) while using significantly fewer credits
Parse Router is now our recommended default for most customers with mixed document payloads in mortgage, insurance, healthcare, and legal
it's live today, set the engine to `parse_auto` and give it a shot on your own docs
Davis Treybig retweeted
Cloudflare, Vercel, Uber, LaunchDarkly, and others are building software factories.
I looked at how they work, the results so far, and what's next: will agents be prompting us?
igoro.com/archive/software-f…
