colincooke

15 ms·

91.
Ocado Acquires Kindred Systems(kindred.ai)
2 points by colincooke 6y ago|0 comments
92.
by colincooke 6y ago
This isn't entirely true. There are many high benefit activities we should target (flights are a good one), however there are also others that are much more difficult to reduce (energy production in emerging economies for example). I t
93.
by colincooke 6y ago
While ideal, this isn't very realistic. Likelier is that we will slowly get off carbon intensive energy sources and production mechanisms, but fail to move off them entirely for a long period. In this period it is essential that we inv
94.
by colincooke 6y ago
This preprint, combined with a recently published study on measurable heart damage [1] makes it clear how little we still know about the long term effects of COVID-19. Similar to this pre-print, [1], looked at both hospitalized and non-hosp
95.
by colincooke 6y ago
I think of it more as an efficiency argument. Does being self aware help with survival? For the sea slug, probably not, for other (higher level) organisms maybe. Is a pareto frontier of awareness? Trading off the increased cognitive ability
96.
by colincooke 6y ago
This is a very important point. Having the flexibility that wealth allows can unlock many creative and ambitious opportunities. I've had a lot of friends talk about ambitious company ideas, giving Tesla/SpaceX as examples. What th
97.
by colincooke 6y ago
The article is an interesting perspective on the factor of choices given in "opinion" polls. Forcing a choice one way or another has a really interesting effect (as in including only agree/disagree, and no "I don't
98.
by colincooke 6y ago
In clock speed perhaps, but in instructions per cycle absolutely not (not even close). So for the vast vast majority of workloads a newer processor will be much better than even an higher clocked but old (E8400 is from 2008) processor.
99.
by colincooke 6y ago
I'm friends with one of the labs that took part in this study, so to give a perspective of how this was received in the field. For the most part it wasn't a radical finding, but really good to quantify and see the effects. It also
100.
by colincooke 6y ago
There is really no strong argument made in this article. It seems like most, if not all, of the problems they talk about are _employee_ problems not slack problems. Slack is simply a medium to chat with your company and team, one which is p
101.
by colincooke 6y ago
The issue with efficient summarization is the necessity of context. For a message to be effectively summarized you require a bit of "theory of mind"[0], you need to have at least a decent idea of what the receiver already knows. T
102.
by colincooke 6y ago
This response confuses me a little bit. Although in their response the Internet Archive refers to the lawsuit attempting to stop controlled digital lending, I believe the lawsuit instead is looking at uncontrolled digital lending (the Emerg
103.
by colincooke 6y ago
Not exactly sure your critique (maybe I'm misunderstanding?). The gist I get from the article is that many Arab states have dependend on oil to stablize their economies/internal politics and fund their way of life, and that while
104.
by colincooke 6y ago
Yeah you're not wrong, but it's a bit misleading. This allows you to run faster, but it does it by allowing you to use a larger batch size (arguably not best practice but your mileage will vary). Memory pooling is a bit different
105.
by colincooke 6y ago
I think the nice Volta cards (V100) does it "properly". But out of reach for most small scale setups (academic labs, prosumer, independent researcher, etc.). Unfortunately the best case for high-mem use-cases is to just rent from
106.
by colincooke 6y ago
Unfortunately multi-GPU training doesn't scale linearly yet [1] so it's often a better call to get a larger card then two smaller ones, at least for the single-model case. [1] https://github.com/keras-team/ker
107.
by colincooke 6y ago
Should note (from someone who has a few of these systems at my lab) unfortunately the consumer RTX cards don't do memory pooling. This means that although NVLINK is good for inter-GPU comms it doesn't actually allow you to run gia
108.
by colincooke 6y ago
I'm curious, as you seem to be well read on this topic, what the effect of desert winds/sand are on panel longevitity and efficiency are?
109.
by colincooke 7y ago
Agreed, also this model fully uses the other images, just not the way that traditional supervised learning would. With "just 10 labels" is more accurate. Impressive results, but this isn't some hyper-convergence technique tha
110.
by colincooke 7y ago
This is a blog not a paper, it seems you wouldn't like the source material: https://arxiv.org/pdf/2001.07685.pdf But you are correct! This way of showing off your work is much nicer than what ends up in the paper.
111.
by colincooke 7y ago
This depends on how well those 10 images are representing the disribution of data for your actual task. With only 10 samples thats highly unlikely. What you are talking about is data augmentation, a strategy we can use to expand our trainin
112.
by colincooke 7y ago
This "debug console" is on networked IP cameras (many of which are open to the web) and available through a hardcoded password. I don't see a convincing argument for malicious intent, more so a dangerous level of incompetence
113.
by colincooke 7y ago
Worth noting that your reference is from 1993, and although your argument still stands for a modern reference: "Around 10% of the world's electricity is generated by about 440 nuclear power reactors" [0] [0] https:/&#x2
114.
by colincooke 7y ago
Completely agreed. I was surpised reading the article about how fusion (irrespective of being able to achieve it with positive energy output) was obviously a non-starter due to neutron embrittlement. I think there is a place for this kind o
115.
by colincooke 7y ago
This is great to see, I've seen lots of other lists on the web but always happy for more (perhaps this will keep up to date). I'd suggest adding a publication date column (should be easy to scrape if need be). One thing I'm u