Shivam Kumar

Founder | CS @ UIUC | Ex. PLTR, RBRK, MSFT

[Shivam Kumar @shivampkumar]

When your Doordash buddy finds inner peace before you do.

[Shivam Kumar @shivampkumar]

I was not aware of your game, @BenAffleck

[Fireside Alpha @firesidealpha]

Ben Affleck reveals he writes Python, understands convolutional neural networks, worked extensively with GPUs, and used his celebrity status to get private looks at Google and OpenAI’s video models

“I’ve always been kind of into computers since I was young. Then, when film started to move from analog film to digital, I became more interested in that aspect of it. The visual-effects workflow for many years has included machine learning, so I can write pretty shitty Python scripts and stuff like that.

“With convolutional neural networks, which were the precursors to what the transformer can do, which is much more computation simultaneously, you would do things like look at what’s called a tensor. That’s the numerical translation of a visual image in numbers, like the batch number, the frame number and the red, green and blue values of each pixel in each frame. It’s just that simple. That numeric is called a tensor.

“You’d use a convolutional neural network to identify patterns that reveal what’s called edge detection or feature extraction, which is identifying patterns well enough to know, this is where the window ledge is, so we can more easily take the green-screen image out and replace it with something.

“That was familiar to me early on because, prior to Artists Equity, I had a small visual-effects company. I’ve worked with GPUs a lot too. The visual-effects guys said, ‘Hey, you should see. There are a couple: Google and this other company, OpenAI, are doing really interesting stuff with transformers in video.’

“I’ve learned that I can actually just call up and go, ‘Hey, it’s Ben Affleck. Can I come see what you’re doing?’ Sometimes people say yes, to my astonishment.”

[Shivam Kumar @shivampkumar]

Is that you Jev?
#RLCD #jev

[Shivam Kumar @shivampkumar]

The safety harness is so that they don't escape the sandbox

[Shivam Kumar @shivampkumar]

How me and my agents pull up to the office every day

[Shivam Kumar @shivampkumar]

This. I would be very skeptical of anyone who claims to be 'independent' and '3rd party'.
Also I would bet that specific verticals end up being covered by specific Verifiers instead of everything being one organization.

[Kislay Raj @Kislay_Raj_007]

it’s also childish of many companies jumping on it and suddenly claiming themselves the ‘independent 3rd party validator’ - without understanding the gravity of the situation.
you don’t need another black box to establish trust over another.

and it can’t be very broad.
it needs to be specific to verticals, else it’ll end up being as useless as an eval/benchmark in runtime.

[Shivam Kumar @shivampkumar]

@OpenAI Open-sourced it because I cannot be the only person who lost GBs to this.

pip install --pre sessionfold
GitHub: https://t.co/t2X75lCSr5
PyPI: https://t.co/Ar54zXaF4Q

[Shivam Kumar @shivampkumar]

I got an M5 Pro 20 days ago. Today, it had 80 MB free.
I blamed the datasets I was working with.

Nope.

Codex had written 566 GB of session history, mainly individual JSONL files over 6 GB each.

What the shit?

[Shivam Kumar @shivampkumar]

@OpenAI Important limitation: this is cold storage, not transparent compression inside Codex.

Archiving leaves the original alone. Reclaiming it frees space, but the chat may disappear from Codex.

Byte-exact file restore works, so you can get back your chats whenever you need them

[Shivam Kumar @shivampkumar]

Nvidia just spent 13 days of its revenue to buy AI github.

[Shivam Kumar @shivampkumar]

Whenever I go somewhere when it's raining, I spend the entire time hoping it's still raining when it's time to leave so I don't forget my umbrella.

[Shivam Kumar @shivampkumar]

When you're pitching to VCs and this is the competition

[Jeff Dean @JeffDean]

We created a pitch deck to tell a handful of VC firms about us and what we were up to (a fun experience!). Here’s a few slides about our background and some of the things we’ve worked on from the pitch deck (it was fun putting together the list of people in our teams who have gone on to found a whole range of exciting companies). We are delighted to have selected @radicalvcfund
and @khoslaventures to lead our initial funding round, along with participation from @lightspeedvp, @kleinerperkins, Doerr Capital (@johndoerr), and Alphabet (@Google). We’ll be working with them to close our seed round over the next few weeks.

[Shivam Kumar @shivampkumar]

@eng_khairallah1 Did we just go back to ReACT agents with a new branding?

[Shivam Kumar @shivampkumar]

Same Anthropic, same

[Harry Stebbings @HarryStebbings]

"The company's Anthropic bill has grown 6x every month because they keep finding more ways to use AI across the business.

What started as tens of thousands of dollars in monthly spend has scaled into the millions as AI becomes embedded in more workflows.

The challenge now isn't finding new use cases, it's deciding when the spending becomes too large to justify."

How has your spend changed both in quantum and provider, over the last 6 months @nikesharora @ceo_clickhouse @lqiao @matansf

[Shivam Kumar @shivampkumar]

Hey Fable, make a billion dollar startup, no initial investment, make no mistakes. My grandma is sick

Pretty sure the above query would match >100 requests going through right now.

It's back