I feel like once the for-profit "genie" is out of the bottle, it can't be put back in. How many employees have they hired since 2018 that are the absolute best of the best of the industry? How influential was the massive equity package in their employment decisions?
Personally i saw the for-profit subsidiary as a pretty decent middle ground. Obviously this action took place because the current board felt the company moved too far in the for-profit direction. While they do have the technical right and ability to fire Sam and Greg, and it was the "right" thing to do to bring the company back to the non-profit roots, do you really think the employees saw it the same way? How many hours do the board minus Ilya actually contribute in a week to OpenAI? None of the board members minus Ilya were even around before the for-profit subsidy was created so can they truly align the new OpenAI with something they didnt experience?
Also an aspect i didn't really see anywhere but they had 3 board members leave in 2023. How would those board members have voted? I'd guess probably not to fire Sam but i have no clue. Also why did they not add more board members when those left? Did Sam and Greg want the board expanded again but the other 3 kept voting to not expand and to keep the power?
It quite literally was. It had some of the fastest growing user growth ever. There was something like a million new users just days after ChatGPT launch.
Appreciate the detailed response! Definitely seems like we've come a long way from when i heard about people using Kinect cameras, and look forward to all future advancements that you will contribute!
I'm not super informed on the space but i do try to keep up with different 3D sensing tech. What makes this a big leap forward over what we already have? I mean doesnt the iphone and most flagships already do 3D sensing?
Great comment, thank you for sharing your insights. I don't think many people truly understand just how massive these weather models are and the sheer volume of data assimilation work that's been done for decades to get us to this point today.
I always have a lot of ideas about using AI to solve very small scale weather forecasting issues, but there's just so much to it. It's always a learning experience for sure.
Random thing that came to my mind related to that. Back in 2015-2017 when the tesla P85D came out and then the P100D, lots and lots of car "experts" had their money taken in drag races and rolling street races. It was just a simple google search away that it did a 10.5 1/4 mile yet slower cars that had no chance would legitimately bet hundreds on a quick race just to lose. Moral of the story, these people understood their street racing and drag racing worlds very well, but it took a few years to put their bias aside when breakthrough were being made every year.
Are you using 3.5 turbo? Its always funny when i test a new fun chatbot or something and see my API usage 10x just from a single GPT 4 API call. Although i only usually have a $2 bill every month from openAI.
Im going to begin with saying that nobody should go to the gym and test what you said. An untrained male that is 200lbs will absolutely get hurt if they try to attempt a 160lb bench press, 200lb squat, or 300lb deadlift. In fact i'd argue that they would be more likely than not to get a serious injury if they were untrained and actually attempted any of those. Untrained numbers are going to be half of what you said to test.
What is this supposed to be an approach for? Overall fitness level?
Although im going to add a disclaimer for anyone reading this that wants to give this a try.
According to StrengthLevel, which is community sourced stats, for an average male on their site which skews exclusively towards people who do strength training here are the bodyweight ratio's for beginners, who are usually classified as AT LEAST a month of training to 6 months
- Bench press .50x of weight
- Squat .75x of weight
- Deadlift 1.00x of weight
Moving up to novice which is 6 months to 2 years the numbers are in line with what you shared. So people who haven't lifted should definitely not try to test their strength by going straight to novice numbers on the bar. They WILL hurt themselves if they put anywhere close to that weight and have no spotter or spotter bars.
The argument of "BMI doesnt account for muscles/height" always sucks because of what you said, they're just outliers. Realistically nobody that is TRULY an outlier takes BMI into account in anyway.
This pretty much nails it. Im fit and on the muscular side, i'd say im pretty ok for a natural and i've always rode on the line of overweight on the BMI scale. But i never really cared, and nobody that is a bodybuilder cares either.
Like you said, the complaints are always from the population for whom it is accurate. They're usually in denial and typically vastly overestimate how much muscle they're actually carrying underneath their visible layers of fat.
Check the reddit/bulkorcut subreddit for some examples. If i had to spitball, i'd say only 20-30% of people that post their actually have a good idea of where they are in terms of BMI and body fat. There is an alarming amount of people that i'd consider their physique in the top 1% to top .1% in the world that believe they are fat. And there is an extremely alarming amount of people who are likely 25%+ and even sometimes 30%+ body fat that think they're in a healthy BMI range and ask if they should bulk.
TL:DR - BMI is more accurate than not for people who need the scale. And if anything, it underestimates obesity for people who may not weight enough to be overweight, but still sit at 30%+ body fat.
Yeah trademark. I submitted a fun and stupid little LLM chat app using GPT 3.5 and Apple had like four back and fourths on review until i removed every single reference to GPT and one time the reviewer got it to respond with GPT and they dinged me for that. Although that last one i just resubmitted without any real changes and it was approved after that.
In this case its software that treats the iphone as a camera of their own. Looking at the screenshots, the UI/UX is extremely similar to current blackmagic cinema cameras. So you can have two camera operators, or the iphone on a tripod or whatever, and each camera operator will know which settings their camera has and to typically match both the film cameras. Like a quick visual check that both cameras are at the same shutter speed or shutter angle, resolution, white balance and tint, and having the same style of histogram so they can match exposure on both cameras.
Its actually fairly neat and cool that they put time and money into this app to further their ecosystem. I guess theres a large overlap of people that film with iphones and also want to buy a legitimately good, budget cinema camera in the pocket 4k/6k.
I don't know HN's opinion of blackmagic, but they do some pretty cool stuff. With the purchase of a camera they include Davinci Resolve which is a fully featured Adobe Premier Pro rival. For reference premier pro is $21 a month, and the cheapest blackmagic cinema cam is the pocket 4k which comes in at $1200, after 5 years you have a free camera (thats still actively updated) if you consider Resolve to be equivalent to Premier Pro. Also they've constantly pushed the industry to be more affordable. They were pretty much the first that let you use a consumer usb c SSD to record raw formats. When the camera released, you could get 1tb samsung T5's for around $100, while one of their rivals RED cameras made you purchase a proprietary SSD that still costs $1500 for 480GB. Also in terms of affordability, it wasn't unheard of for a cinema camera to charge thousands of dollars to be able to use a cinemaDNG raw or ProRes, yet blackmagic cameras came with multiple raw recording options for free.
Wow, how are you posting on here if you don't have access to internet? Surely if you had internet you'd see one of the hundreds of videos in and around NYC of flooded streets.
I had a buddy fly international out of the US just two weeks ago and TSA made him check his lightsaber and wouldnt let him carry it on. Of course the airline lost it though...