smeeth

17 ms·

151.
by smeeth 6y ago
They were in 1987 when SBIR gave them their first grant [0]. That's sort of the point, parent comment asked for an example of a company that exited with SBIR help, and I gave him a wildly successful one. [0] https://www.sbir
152.
by smeeth 6y ago
I'm conflicted. On one hand, I don't believe citizens should have the ability to protect communications (or anything, really) from a lawful warrant. It is obvious that the right to privacy is not a right to perfect forward secrecy
153.
by smeeth 6y ago
“This funding allowed us to pursue several innovative programs that otherwise would not have been possible” - Irwin Jacobs, founder of Qualcomm in testimony to the US senate. The government is good at a lot of things, it is NOT good at publ
154.
by smeeth 6y ago
With great power comes great responsibility, Spotify Data Scientists. If discover weekly is truly that impactful a decision made by a model to include or not include a song could determine whether or not an indie artist can pay their rent t
155.
by smeeth 6y ago
> As long as they are printing true facts, they can only be doing good This is obviously not true. Releasing a fact has a tangible effect on the world, usually a slight change of opinion in the minds of those who learn the fact. Someti
156.
by smeeth 6y ago
Step 1: create a repo with 4K stars...
157.
by smeeth 6y ago
This is a bit like telling the whole CS community in the 90s that we need a year off the internet to study ethics because people can use it for child porn, money laundering, and cybercrime. Should we have?
158.
by smeeth 6y ago
Feed structure and content aside, is writing and applying a custom TOS that covers content not editorializing/publishing? This is the crux of the argument from the political right at the moment: Twitter is deciding what sort of content
159.
by smeeth 6y ago
That sounds to me like the outcome of both options are: 1) Toxic, awful to use sites with no personalization (custom recommendations, curation, etc) 2) Sites get more careful (and expensive to operate) because they don't want to get su
160.
by smeeth 6y ago
Of course this is being handled ham-handedly, but it seems to me that section 230 is a ludicrous piece of legislative malfeasance that allows content publishers to duck the legal responsibilities of publishing while profiting from behavior
161.
by smeeth 6y ago
Very much agree about old NASA suits - especially the orange shuttle ones. I think it's important to discriminate between what the spacex suits look like in stylized prototypes and actually on people. Take this image for example: http
162.
by smeeth 6y ago
Right? Reading some of the comments here I thought I was going crazy - the suits look like they were made for a 2003 made-for-tv movie. Boeing suits look WAY better imo
163.
by smeeth 6y ago
I’m a simple man, I see an N=1 experiment, I ignore it.
164.
by smeeth 6y ago
As someone who uses pandas daily and detests it violently, godspeed.
165.
by smeeth 6y ago
I really find it difficult to put into words just how little I care to pay for a web ui so I can "manage" my data. Data pipelines are a real problem though, and I'm very interested in what startups do with this space.
166.
by smeeth 6y ago
Very useful! Thanks! > Although PyTorch has offered a series of tutorials on distributed training, I found it insufficient or overwhelming to help the beginners to do state-of-the-art PyTorch distributed training Can confirm. Am relative
167.
by smeeth 6y ago
Intelligence Matters: foreign affairs podcast hosted by former #2 at the CIA. He’s a pretty good interviewer and I’ve learned a ton about how our govt perceives global threats. Pivot: Kara Swisher and Scott Galloway talk (mostly tech) busin
168.
by smeeth 7y ago
I'd be best described as an ML Researcher/Engineer and I'm not in the private sector so take my opinion with a grain of salt, but my understanding is that many DS roles require application of the scientific method. A lot of D
169.
by smeeth 7y ago
I always thought the non-specificity of the term Data Science was a strange criticism for those in the tech industry to make. How many types of SWE are there? Front-end, back-end, full-stack, devops, security, QA... I agree wholeheartedly
170.
by smeeth 7y ago
Like so many articles on ML the details are sound and the "so what?" is off. Literally the first line of this post is wrong/misleading; CNNs can and have be used to process temporal or sequential data (1d CNNs in particular).
171.
by smeeth 7y ago
My primary work use of search is looking for academic papers. I understand its a rarer use-case but DDG just isn't there yet unfortunately. Looking forward to when they are!
172.
by smeeth 7y ago
Apparently Bayer was able to manufacture and ship an order of 300,000 chloroquine phosphate pills in 24 hours. I don't know exactly what that means, but it's definitely good. https://www.shine.cn/biz/economy&#
173.
by smeeth 7y ago
I think, though I'm not sure, that RH makes their money by investing in treasury bills with the cash balances of people's portfolios. You're (most likely) not getting secretly screwed.
174.
by smeeth 7y ago
It really depends on the specifics of the optimization problem in question. If ML is trained on biased data where an optimality exists only at an unbiased solution (think a shaped reward function in RL to disincentivize class bias or someth