Ramesh (CV) Chitor

Making #AI Safe & Trustworthy #FTTB Investor, Advisor @perplexity #trustmodelai #rubrikinc
#jvccard #spacex

[Ramesh (CV) Chitor @rameshchitor]

It was a rare honor meeting Honarable Governor of Telangana, Shri Jishnu Dev Sharma at Lok Bhavan, Hyderabad today. AerovaCare is dedicated to evangelize and support in ending #Tuberculosis and other airborne diseases. @PMOIndia @TelanganaCMO @TelanganaDGP @TelanganaDGP @TelanganaGov @Indianredcross2

[Ramesh (CV) Chitor @rameshchitor]

@rwang0 @FIFAWorldCup Great PoV

[Sundar Pichai @sundarpichai]

Our new Gemma 4 12B model hits a sweet spot between size + performance: it can run locally on a laptop, while enabling powerful multi-step reasoning and agentic workflows. Can’t wait to see what the community does with this one!

[Karl Mehta @karlmehta]

Geoffrey Hinton says we're probably not slowing AI down.

So the real question is:

can we make it safe before it reaches customers, employees, and workflows?

Karl's article below turns the warning into the builder question:

what did you test before you shipped?

[Bipul Sinha @bipulsinha]

Rubrik Cyber Resilience!

[NVIDIA @nvidia]

Taipei runs on innovation and great food. 🥟🍧

If you're at GTC Taipei at Computex, visiting the night markets is a non-negotiable.

#NVIDIAGTC

[Karl Mehta @karlmehta]

This isn't hypothetical:

• Mobley v. Workday → AI hiring bias, nationwide collective action
• Air Canada → liable for what its chatbot said
• EU AI Act → fines up to 7% of global revenue

"We tested it" isn't a defense. "Here are the scores" is.

[Karl Mehta @karlmehta]

Enterprises are carrying ~$500B in *unvalidated* AI risk.
100+ AI lawsuits are already active in US courts.

Most AI still ships without anyone checking if it's safe, fair, or defensible.

So today I'm open-sourcing TrustModel: 🧵

[Karl Mehta @karlmehta]

https://t.co/Ew9NbCovBX

[Konstantin Kisin @KonstantinKisin]

There is something darkly amusing about the fact that selling victimhood to the most privileged people in history has become such a lucrative and big business.

When I was on tour with @jordanbpeterson he talked about many things, but probably the most common recurring theme was the "Spirit of Cain". It seems our ancient and sacred texts tell these stories for a reason: victimhood is easy, seductive and addictive. And now profitable too.

We are living through a perpetual victimhood escalation battle where people (and groups) now compete not on merit, but on the supposed disadvantages they face. Which makes perfect sense since this is the incentive structure our societies have been encouraged and forced to adopt.

[Ramesh (CV) Chitor @rameshchitor]

https://t.co/QORFnQQKZh We have now open sourced our fix! at @trustmodelai A huge step making https://t.co/puJvsBUSdg part of the open source community. Run it locally, read the code and score your AI!. Why do you need a sales call, anymore? Brilliant @karlmehta

[Ramesh (CV) Chitor @rameshchitor]

The question isn't whether your AI works it's whether you can prove it works to your regulator, your board, and your customers.

CEOs: Stop bragging on how cool your AI platform or skills are. Get to the meat and potatoes of problem solving and impact.

Get US engaged @trustmodelai to keep YOU safe!

[Ramesh (CV) Chitor @rameshchitor]

The question isn't whether your AI works it's whether you can prove it works to your regulator, your board, and your customers.

CEOs: Stop bragging on how cool your AI platform or skills are. Get to the meat and potatoes of problem solving and impact.

Get US engaged @trustmodelai to keep YOU safe!

[Ramesh (CV) Chitor @rameshchitor]

@_mohansolo Exciting to see what you are building at Antigravity

[Ramesh (CV) Chitor @rameshchitor]

@GeminiApp #Google you really messed with me during the drive to work reporting incorrectly on a Live #iplseason2026 game. Never expected this out of you 😭. Playing with fans emotions does not bode well.

I'm taking it in a sportive spirit. Now, I know you can mess with me!

[Ramesh (CV) Chitor @rameshchitor]

The question isn't whether your AI works it's whether you can prove it works to your regulator, your board, and your customers.

CEOs: Stop bragging on how cool your AI platform or skills are. Get to the meat and potatoes of problem solving and impact.

Get US engaged @trustmodelai to keep YOU safe!

[Sarbjeet Johal @sarbjeetjohal]

A beautiful day comes to an end in Silicon Valley!

[Ramesh (CV) Chitor @rameshchitor]

@karpathy massive!

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[Karl Mehta @karlmehta]

Karpathy's Software 3.0 framing is the missing backdrop.

If prompts are programs, then context, tools, memory, evals, and permissions become infrastructure.

That is why production agents need engineering discipline, not prompt theatrics.

[Ramesh (CV) Chitor @rameshchitor]

@karlmehta exciting time to build

[Karl Mehta @karlmehta]

https://t.co/WXNNGP8jJJ