Jayanth Janardhan

Product manager - Meta

[Jayanth Janardhan @JayanthJanardha]

I definitely thought this video was AI generated. Very few people actually understanding the inner working of models to then accurately understand and apply to their domains and use case. Very refreshing from @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.”

[Jayanth Janardhan @JayanthJanardha]

@OpenAIDevs Feedback on the CX: Agent CX have progressed from human-in-the-loop to human-on-the-loop experiences. Consumers expect agents to just "do stuff" on their behalf either proactively or without constant back and forth. Appreciate the re-take tho. Thank you.

[Jayanth Janardhan @JayanthJanardha]

@OpenAIDevs Hugely appreciate the re-take. Being a product person myself, I applaude the courage it takes for someone to be “in the arena”. Haters gonna hate.

[Jayanth Janardhan @JayanthJanardha]

In the era of personal agents, Apple is the most under rated and likely the biggest beneficiary. Personal agents might just revive apple.
#apple #muse #agents

[Jayanth Janardhan @JayanthJanardha]

@HedgeyeComm You forgot the biggest of them all - $AMZN

[Jayanth Janardhan @JayanthJanardha]

@viticci Why would you do this instead of using Claude code directly?

[Jayanth Janardhan @JayanthJanardha]

@FredaDuan @openclaw Enterprise agents (the one similar to Autopilot) is way more stickier than a consumer agents like Instinct/Muse. Enterprise has a lot more use cases (stay on top of mails, chat threads, compose weekly status updates, write docs, presentations, demos etc.) than consumer agents

[Jayanth Janardhan @JayanthJanardha]

@antigravity You guys seem to be at least 6 months behind the industry leading coding agents

[Jayanth Janardhan @JayanthJanardha]

@RihardJarc None of these agents being built by itself is defensible. We will have 100s of companies that will clone openclaw to build their own version of Muse.
Enterprise will see stronger adoption with Microsoft/AWS winning but consumer space will be a lot more fragmented

[Jayanth Janardhan @JayanthJanardha]

3) Find the sticky CX. Initial traction is great, but novelty fades. Once people have “played” with Muse for a few weeks, what brings them back every day?

[Jayanth Janardhan @JayanthJanardha]

First off, congrats to the Muse team — fantastic launch. And the mascot is 🔥

Now comes the harder part: how does Muse turn launch momentum into a durable business?

Three things I’d be thinking about:

[Jayanth Janardhan @JayanthJanardha]

2) The product is clonable. If the experience works, competitors can reverse-engineer the core ideas relatively quickly. The moat has to become more than the product itself.

[Jayanth Janardhan @JayanthJanardha]

1) Distribution is king. OpenAI and Google are likely thinking about similar experiences, but Apple may be the dark horse given its massive device distribution.

[Jayanth Janardhan @JayanthJanardha]

@thdxr Not true. Azure is the only hyper scaler that forces you to do provisioned throughput. Both GCO and AWS offer on demand for large amounts of compute

[Jayanth Janardhan @JayanthJanardha]

@L_BARROS @WSJ @grok @grok does this apply to applicants converting from temporary to permanent green card?

[Jayanth Janardhan @JayanthJanardha]

@thsottiaux what benchmarks do you use for your evals?

[Jayanth Janardhan @JayanthJanardha]

@amanrsanger Did they train a new base model ground up? Or did they fine tune an OSS model?

[Jayanth Janardhan @JayanthJanardha]

@trq212 How do I sign up? And what’s the bridge?

[Jayanth Janardhan @JayanthJanardha]

@theemozilla The whole team that built Llama moved on to start Mistral. Meta lost a lot of AI research talent

[Jayanth Janardhan @JayanthJanardha]

@GerberKawasaki It looks like shit. It’s never going to be a mass market truck