1,481 Photos and videos
21h
Wow.
USG needs model independent infra. As does every business of consequence. H/t @capitolai
JUST IN: Pentagon says more than 2/3 of daily AI workflows have been shifted off Anthropic to competing AI vendors
1
139
Jun 12
As AI becomes embedded in everyday work, the challenge shifts from access to control.
I joined @AmirBormand on @TheTechTrek podcast to discuss this challenge and dive into data sovereignty, shadow AI, and why disconnected AI workflows can create fragmentation, governance complexity, and risk inside organizations.
Thanks for the conversation, Amir.
youtu.be/lFuvTWOQRuM
2
54
Shaun Modi retweeted
Jun 10
Our Anthropic bill is about to jump from $400K ā $1.4M/yr.
Not because usage exploded, but because we're about to cross 150 seats.
Past 150 seats you're forced into Enterprise tier. Seats stop including any usage, every token bills at standard API rates. At our current run rate that's 3.5x overnight.
Unfiltered thoughts on AI spend:
1. We should spend tokens to grow as aggressively as possible. But most people (me included) aren't conscious of what they're spending.
2. Visibility comes first. People see their personal number and they're shocked. I accidentally spent $4,000 in 3 days in Claude Code.
3. For engineering the spend is clearly worth it. Pay for the best model, it saves more than it costs.
4. For a lot of other roles it's questionable. Apps nobody uses, skills someone already built. No ROI.
5. Spend limits are coming. We already require approval for more tokens on our support team.
The era of token-maxxing is coming to an end.
367
163
3,022
1,786,977
Jun 9
What a milestone week across the pond for @capitolai presenting our technology at the London Stock Exchange, meeting with customers, and MPs. It was a rare privilege to be invited to 10 Downing Street to meet with the Prime Ministerās Chief Business, Investment and Trade Advisor to speak about Capitolās growth and investment in the United Kingdom / Europe. Hereās to the special relationship. Big things ahead! š¬š§š¤šŗšø @LSEplc @10DowningStreet
1
77
Jun 7
Wow
Jun 7
This is a pretty striking shift toward Chinese models by American AI startups since the start of the year. substack.com/@profgmarkets/pā¦
88
Jun 7
Have this problem? Use @capitolai
Jun 7
Anthropic's Opus 4.7 and 4.8 models are experiencing degraded performance, which is causing a higher rate of failures for users selecting these models in Notion AI.
To mitigate impact, all Anthropic models have been disabled in the model picker and requests have been rerouted to alternative providers. Most users should now be able to continue using Notion AI with minimal disruption, though Anthropic-specific features remain unavailable.
Please refer to notion-status.com/ for the details.
1
163
Jun 2
Every company, government agency, and institution is navigating the same dynamic: more information, more complexity, and less time to make sense of it.
I joined @mattlevenhagen on The Builders Podcast to talk about that challenge, the journey to founding @CapitolAI, and why we believe the next phase of AI is about helping expertise scale.
Appreciate the thoughtful conversation, Matt.
Listen here: youtube.com/watch?v=YoFKwjdSā¦
1
102
Jun 1
A lot of AI products today remind me of Nokia in 2006. Technically impressive. Market-leading, even. And completely missing the point.
Back then, every phone had a different interface and the burden was on the user to adapt to the machine.
Then the iPhone came and shifted the landscape with one universal interface. Software was now part of the user experience.
AI today is still in the āevery device has a different keyboardā phase. Different prompts. Different behaviors. Different levels of reliability depending on how you ask.
Weāre asking users to learn the systems instead of designing the systems around them.
Thatās not a model problem. Itās a design failure.
That idea has followed me from Nokia and NASA to Airbnb and now Capitol AI.
Weāre not just building more AI. Weāre building systems that help people make sense of complexity without forcing them to become AI experts first.
Read more about this in our latest blog here: capitol.ai/blog/why-aiāā¦
1
106
Jun 1
In a world where AI is everything, itās important to take a step back and think about history. Had a special time visiting Bentonville Arkansas to visit the first Walmart, and @crystalbridges, a PHENOMENAL museum of American art
1
157
May 31
šÆ
This is the actual bottleneck. The models are smart enough already. What is missing is the company-specific context locked in senior people heads. Whoever cracks knowledge extraction at the company level unlocks the rest.
As you work on this, please consider using GBrain as your OSS retrieval layer
x.com/t_blom/status/20608063ā¦
119
Shaun Modi retweeted
May 25
Grateful today as we honor and remember the brave.
1
2
66
May 25
Nothing beats culture and community š®šŖš®š³. (And this is coming from an AI guy)
2
122
May 22
May 22
Microsoft is letting Office users remove an annoying Copilot button. A floating Copilot button has irritated Excel users the most, and some relief is coming next week. Details š theverge.com/news/935967/micā¦
161
May 20
Every era of technology has its iPhone moment. The moment a category stops being for specialists and becomes something for everyone.
The personal computer. The smartphone. The internet itself. Each was built by engineers, for engineers, until someone designed it for everyone else. Today we are at that moment in AI.
I've spent my career at the intersection of complexity and clarity. At NASA, I worked on systems where the stakes of a bad interface weren't frustration, they were catastrophic failure. At Airbnb, I learned that trust, at scale, is a design problem. When millions of strangers are asked to open their homes to each other, the product has to feel safe and natural, or the whole thing collapses.
There's no engineering shortcut to human trust. It has to be designed into the system itself.
That design lens is what I brought to @capitolai. And it's why I believe the next great competitive unlock in artificial intelligence is in the product design rather than the model.
I wrote about this on our blog. Read it here: capitol.ai/blog/why-aiāā¦
2
103
