@aprildunford we built a proxy from your positioning frameworks and asked it to diagnose a startup’s messaging problem.
The diagnosis is brutally direct — and I suspect you’d disagree with one key part. You can ask yourself here:
https://t.co/4sVN3lcPew
Hey @EvanKirstel! We made an AI Proxy of you. It's pretty bang on - people could actually talk to it instead of you.
If you've wanted a chat with Evan, you can have one now for free! #TryDotted
https://t.co/Cx31UQeHyY
@alexstamos, we built a proxy from your public security work and asked it how a major platform breach would actually unfold today. The answer is… uncomfortable.
https://t.co/Vu1zX1mPqj
@balajis we built a proxy from your public essays and threads and asked it to make one concrete prediction about the next phase of “network state” ideas.
It gave a surprisingly specific answer — we can’t decide if it’s insightful or completely wrong.
https://t.co/tjNd8ek22V
@levie, we built a proxy from your public SaaS and AI threads and forced it to answer one question: what enterprise software category AI will quietly destroy first.
It picked one and explained why. Check it out:
https://t.co/1gQ2tW4mZU
Our AI Proxy of @AndrewYNg doesn’t do AI hype. It does: pick a use case, fix the data, ship a v.1, iterate.
Ask it your “where do I start?” question: https://t.co/Lzw2zZNL1a
If your security plan can’t survive a chat with our AI Proxy of @schneierblog, it’s not a plan—it’s a vibe.
Stress-test it here: https://t.co/CVOFww1yFp
#cybersecurity#privacy#risk
But here's what most people miss: you can start with the free tier using your own data, on your own time, before involving anyone else.
No procurement process required to test.
"I'd have to get infosec to approve this." That's fine, good even.
Here's what we tell security teams:
→ SOC2 compliant
→ We don't train models on your data
→ Can be configured as read-only access to your tools (Jira, Linear, GitHub)
→ Trusted by teams in the Fortune 500