Gary Marcus

As I said, they did not just use one object -- they used five progressively more general cubes over the course of the project. Not all of them have sensors. If they had hidden the limitations in the paper, I'd agree with adjusting title/abstract. But it's all in the blog post.

yes i know about the five variants on a cube; it’s still essentially one object. i responded to this elsewhere. since you still have neither addressed my key points nor apologized for inaccurately claiming i made an error on the sensors, we are done.

as postscript watch the openai video: NOTHING that i say on my slide is mentioned there. no mention of sensors; no mention of dropped cubes; no mention of innate components. as a future PhD, do you think that is ok?

It's the banner of my twitter profile. As for where I am coming from, I am no longer on OpenAI's payroll and have no current collaborations with anyone there.

I understand your concern for hype in the industry, but OpenAI is not a culprit. I spent six month there and experienced nothing but genuine passion for advancing AI from everyone there.

i am sure there is genuine passion, but there was also overselling of GPT2, somewhat misleading claims about DOTA etc. the excellent research would in the end be better served by toned-down PR. many people have started to feel burned; i gave them voice, but I am not alone.

I respectfully disagree -- at most, you could call their blog somewhat high energy, but why shouldn't it be? Writing formal manuscripts gets dull, and they provide plenty of those anyways. Blogs let them share their ideas in a less constrained form every once in a while.