hadsed

15 ms·

61.
Why would data visualization die? AMA with author of D3.js(reddit.com)
4 points by hadsed 7y ago|0 comments
62.
Itables: Interactive Tables in Jupyter(github.com)
3 points by hadsed 7y ago|0 comments
63.
by hadsed 7y ago
Your second point seems really overblown. Language evolves, embrace and accept that. People seem to be using the words humanities and liberal arts to mean the same thing these days.
64.
by hadsed 7y ago
Heh. I'm not sure you're wrong to say it either way.
65.
by hadsed 7y ago
Can you elaborate on what you think might make it easier?
66.
by hadsed 7y ago
Without really digging into this, if it's a neural model like a sibling comment mentions perhaps you can do model compression that shrinks many multiples, if that is the majority of the image size.
67.
by hadsed 7y ago
Think of it like discovering a new human language. We use logic along with knowledge about the world to triangulate words that deal with concepts. ML models work in the same way. Difference is they aren't human so it's harder to j
68.
by hadsed 7y ago
I always tend to get interested in some technology or library and go straight to the docs. They usually have intro or tutorial/gallery type sections that help you get started. It helps to have a use case in mind and imagine applying it
69.
by hadsed 7y ago
Nice blog post! Certainly puts things into perspective in terms of how one should deal with errors, including sometimes just not caring about them much.
70.
by hadsed 7y ago
What's the difference between a high end Chromebook and, say, a MacBook Air? The latter of which is used by some (though certainly not most) technical professionals.
71.
by hadsed 7y ago
I think it could be done. My thought is that it needs to be well integrated with cloud tools where you can organize and run your data processing jobs for it to be really compelling. The transition between crummy dev work and turning it into
72.
by hadsed 7y ago
Comet looks great, thanks for the suggestion.
73.
by hadsed 7y ago
I fully agree, but the tooling is super immature. I think there's going to be incredible opportunity for engineers to build tools for doing ML in a very efficient and scientifically rigorous way. For instance, Jupyter is the best thing
74.
by hadsed 7y ago
Is this not the assumption of many SQL databases that make certain guarantees about complying with ACID? I'm not a DB expert but feels like something I've read before.
75.
by hadsed 7y ago
So your watch also has a 4G connection? Do you have to pay for that separately through your cell provider?
76.
by hadsed 7y ago
I think a lot of responses to this are missing something unsaid here: a single interview day for someone doesn't always give them a chance to be their best selves. That's ok but you do what you can given this fact. The whole thing
77.
by hadsed 7y ago
Doesn't $1bn seem excessive for being wrongly accused of petty theft on multiple counts? I get that it's a very disruptive thing to have happen to you, but still. Also, the case seems like it has nothing to do with the case except
78.
by hadsed 7y ago
This seems excessive. Prisons don't benefit you the same way your national or state governments do. It sounds like you're saying any kind of government or societal control is equivalent to control exerted by prisons on the prisone
79.
by hadsed 7y ago
Very much agree. To expand on this, check out Stuart Russell and Peter Norvigs intro book on AI. It supports your comment as there's an entire section (chapter?) on path planning and the like.
80.
by hadsed 7y ago
Can you elaborate on why downloading from S3 at startup is a bad idea? And why not synchronous everywhere as opposed to always queues? Good points overall that I'd agree with.
81.
by hadsed 7y ago
I think you're making a reasonable point that's coming in too strong. - I've been in several companies whose teams were not white male dominated, including leadership. - The situation with women in engineering teams is a prob
82.
by hadsed 8y ago
It shows the input output pairs you would use to train the network. Projection is simply your fully connected layer of dimension the embedding size you want (e.g., something like 300). The output column is what is being predicted by the mod
83.
by hadsed 8y ago
Say what you want about Facebook, but at least they put an incredible amount of resources into hiring humans to deal with their problems that require good human judgment. They can of course do this because they basically print money. What a
84.
by hadsed 8y ago
Same. A fantastic book.
85.
How Will We Prevent AI-Based Forgery? (HBR)(hbr.org)
3 points by hadsed 8y ago|0 comments
86.
TradeLens: Blockchain for Shipping by Maersk and IBM(supplychaindive.com)
1 points by hadsed 8y ago|0 comments
87.
by hadsed 8y ago
I think this is a great comment. The most critical aspect is understanding the risks and mitigating them (which is your question about how to get set up for success).
88.
by hadsed 8y ago
The assumption when people say X is the problem is that it's the ROOT problem. So if you take it that way then everything is consistent including this author's title.
89.
by hadsed 8y ago
Is this a problem in AR (as opposed to VR)?
90.
by hadsed 8y ago
Well the medium partner program is meant to enable people to earn money from their writing, no? As an author you have to make a choice to do that. I don't see a problem with this and if anything it's a great innovation. My problem

More ›