sytelus

53 ms·

691.
by sytelus 6y ago
Availability accelerates the process by an order of magnitude.
692.
by sytelus 6y ago
That's not how it works. I think you have little idea, if any, how easy it is to find exploits if you had a source code vs if you didn't. With source code you can immediately see the conditions that will cause buffer overruns, for
693.
by sytelus 6y ago
Good investors don't rely on P/E ratios. At this point, GAAP earnings are basically becoming smoke and mirrors: https://www.forbes.com/sites/greatspeculations/2018/11/19/pe...
694.
by sytelus 6y ago
I'd read about an article how there is a significant "dark" black hat companies which do nothing but dig up into the open source codes to find the exploits. They invest big time to develop proprietary tools for autonomous sca
695.
by sytelus 6y ago
Typically insurance payment doesn't cover everything. You will almost always be at some loss, often much greater than year worth of rent. Also, I'd suspect your future premium rates would rise given insurance companies will see in
696.
by sytelus 6y ago
Meanwhile stock market is preparing to for its record high :).
697.
by sytelus 6y ago
Most definitely NOT the advertisers who wants to pay the biggest bugs or yielding huge ad click throughs. I would expect Google ranking reviews from more honest sites like consumer reports, BBB, chiropractor reviews etc and supported by art
698.
by sytelus 6y ago
No. Mattress business is one giant hornet's nest where deception, false claims and high margin are norm. Most companies who sale mattress actually don't do squat of real research or improvement or even own design but they keep cla
699.
by sytelus 6y ago
Classic story of how “activists investor” like Carl Icahn are sucking out blood out of perfectly functional companies. You find vulnerable companies, load them up with debt, transfer their assets out and then have them declare bankruptcy.
700.
by sytelus 6y ago
I have lived by quote “change is better” and for jobs it has almost always worked out because fortunately there are plenty of equal or better jobs in IT. But I am not sure this would work out for an aeronautical engineer at Boeing in same w
701.
by sytelus 6y ago
Why this should depend on convexity at all? The study’s premise is marginal benefit, i.e., estimated risk difference is either minimal or perhaps too poorly known to call it out as marginal. If it was known to be positive or negative, actio
702.
by sytelus 6y ago
In principle, it is fair to get paid same regardless of where you live because you should be getting paid for your contributions, not where you happen to connect your laptop. However, in current form of capitalism, you don’t get paid in p
703.
by sytelus 6y ago
Let's don't forget they have pulled this off during COVID times when supply chains are under pressure, offices are closed, gathering is hard... If you can put man in space during COVID, you can do anything. Hats off to you, again,
704.
by sytelus 6y ago
It’s absolutely not. I don’t follow anything political, neither Obama nor Trump. Among 1200+ people I follow are all from AI/ML/biotech/space research. Most of them are academics. I throughly enjoy my feed and it’s my go to p
705.
by sytelus 6y ago
I’m now fully expecting republican mega donors coming together to trying to gain control of Tweeter as strategic asset. At $23B market cap it’s fairly easy target given there are already other major investors wanting to see Dorsey out. Ther
706.
by sytelus 6y ago
You cannot not have bias when you are talking politics. For Tweeter’s case, one would need to look at statistics for how many tweets from each sides are getting flagged. If one side is getting all the flake then it means that side is evil a
707.
by sytelus 6y ago
This doesn’t work for political tweets. Look at replies to even Trumps benign tweets and you will see 50% of the population would hate other guy no matter what they tweet. Every single tweet of Joe Biden and Trump will get flagged no matter
708.
by sytelus 6y ago
There are 145M daily active users on Tweeter. So that’s approx 3 tweets per user. Sounds reasonable.
709.
by sytelus 6y ago
How did they got #1 and #2 results above real website which even has shorter domain name? My guess: They bought off fake URL in various yellow pages listing which Google then aggregates to believe it is the official URL.
710.
by sytelus 6y ago
In US liability is everywhere and in everything. With that logic it would be impossible to sell anything (imagine liability associated with a copy of Windows!). I'm not lawyer but I can imagine donating these bikes with clear legal con
711.
by sytelus 6y ago
There are two kinds of idea generation: hole plugging and door opening. The former means you look for holes in current state of affairs and use available tech to plug it. The later means you invent something very new and open doors to entir
712.
by sytelus 6y ago
GPT-3/175B model required 3.14E23 flops of compute for training. Even at theoretical 28 TFLOPS for V100 and lowest reserved Azure pricing, this will take 355 GPU-years and cost $3.6M for a single training run!
713.
by sytelus 6y ago
Why selling these is so hard? There is a huge market for second hand bikes. I would think you can always get a better price than scrap metal for a functional bike. Any non zero profit is better than zero.
714.
by sytelus 6y ago
This is actually much more far reaching. Websites modify user content all the time. For example, if you post link on Twitter, you don't decide thumbnails. Another example, appending an ad or "fact check" link are modification
715.
by sytelus 6y ago
Not a
716.
by sytelus 6y ago
Another relevant paper is Intelligence Without Representation https://people.csail.mit.edu/brooks/papers/representation.pd...
717.
by sytelus 6y ago
We been hearing about Threadripper a lot. How well this works for Windows? Does it make CPU bound ML/DL/data science computation better? Any good benchmarks in this area to look at?
718.
by sytelus 6y ago
Have you fact checked yourself? * On 30 April 1905, Einstein completed his thesis* https://en.m.wikipedia.org/wiki/Albert_Einstein#First_scient...
719.
by sytelus 6y ago
No, Eienstein got his PhD in 1905 and he couldn’t find academic jobs, partly because of being Jew and partly because he had burned bridges with few professors, which is why he had to work at patent office. His special relativity paper was p
720.
by sytelus 6y ago
One issue is that reviewers themselves don’t get rated. They also don’t have any incentives to do quality reviews. I think StackOverflow model is the only one I know of which is able to scale this process. It’s not without flaws but it’s co

More ›