Its been so many years since I heard about BMVC. Almost surprised its still around.

I like the phrase, "Don't try to be humble. You're not that great!"

If you're Torralba or Winston, it's okay to say "thank you".

If you're still rising up the career ladder, it's better to make those subtle power moves. Most MBA programs have an Executive Presence course which agree w/ Winston.

Looks like you're the favorite child!

I never hear humanoid robots startups talking about stereo matching or depth estimation or lidar, which I find a bit odd. Maybe it's cooler to talk about VLMs?

With more young researchers in the community, I am not surprised. When we benefit from young and passionate people contributing to advancing our science, I think this is a small price to pay. Seniors should just forgive and forget, have more workshops where the process is made transparent.

Every single shot of teleoperated robots needs to have a clearly visible “TELEOPERATED” label in the corner! Teleoperated demos are great to show off your impressive hardware but it’s grossly misleading to let people think they’re watching an autonomous robot, driven by AI, when they aren’t.

Looks like plain "daal chawal" to me.

My vote is for hype. He needs to say this to appease his shareholders. I expect 10-20% productivity boosts but Meta is sitting on piles of cash and should be building in many directions simultaneously. Lack of human resources has been their biggest blocker, so they can't stop hiring in the long run.

Over the past 6 years, I've gotten decent at "real business" (as opposed to lavishly spending VC money). I raised a couple of relatively small financing rounds and my company works with a dozen large manufacturers who provide us with real revenues that feed our operations.

My current interests are in video activity recognition in unsupervised and multi-modal settings. I work with a small team (5ish) of junior researchers who publish and a 10x larger engineering team who build those capabilities into products.

Followed by a postdoc at Imperial College London on looking at Visual SLAM from a hardware/software optimization perspective, and stints as a Research Scientist at industrial research labs, Qualcomm, NEC, Microsoft.

Most fresh MSc-holders who get a job at Google will get a much weaker direct boss than if they went to CTU or TUM or ETH.

Whether you love your boss or hate your boss, at that age, you absorb a lot from your boss and those behaviors stay with you for life. So it's generally better to opt for PhD.

One nuance missed in this discussion is the caliber of the PhD advisor versus manager at a company.

A PhD advisor at a mid-tier EU/US University is far smarter than an SDE Manager at a mid-tier company; and a professor from a top-tier university is VP-level when she goes big tech.

So far those promises have been kept, if you discount the naysayers who are always moving the goalposts. I am an AI researcher too and compared to the pace of scientific advancements in the first 10 years of my career, i.e. 2007-2017, the rate of concrete annual advances now is a 100x.

Steven, you seem to be living in a bubble. Go talk to the 50% population with <100 IQ.

These tools are advertised as being smart Copilots for day to day tasks. Most normal people dont do physics, mathematics, or complex reasoning tasks.