fisf

6 ms·

31.
by fisf 2y ago
Equity is one thing. What really matters is voting rights.
32.
by fisf 2y ago
Yes. And liberal democracy (the US in particular) is far from perfect. But it is objectively better than a one party dictatorship. It is also less "threatening" from an international relations perspective.
33.
by fisf 2y ago
To stop being a one party dictatorship would be a start.
34.
by fisf 2y ago
Yes, but that is a choice. It doesn't have to do that. Because, instead of "trusting" the update (or rather codebase) of a messenger, we now have to trust the complete browser bundle.
35.
by fisf 2y ago
So, companies like the free market when it suits them, but want regional monopolies (without providing any value) when it benefits the consumer. Interesting.
36.
by fisf 2y ago
GDB can do a, b, c (async Code is not really special?). d is going to be tricky and really depends.
37.
by fisf 2y ago
People who think that case X cannot be debugged without printf often don't know the features of their debugger. I.e. look at several of the comments which seem to miss that you can: - Remote debug. - Use conditional breakpoints. - Use
38.
by fisf 2y ago
Easy. Runtime / performance. Peak memory consumption. Correct behaviour for injected faults, latency, or invalid input. Fault free operation under simulation. The list goes on.
39.
by fisf 2y ago
Because they have not understood the core issue. It's impossible to store / sanitize data correctly, when this is absolutely context / output dependent.
40.
by fisf 2y ago
Because you can enforce password policies without the password ever leaving the (untrusted) client in clear text. I.e. the server only sees the hash and still knows it's dealing with a strong PW.
41.
by fisf 2y ago
If the result of political correctness is an inability to point out competency problems (for whatever reasons), then we have major issues.
42.
by fisf 2y ago
Yes. And to spell it out explicitly: this is yet another instance of a big company willingly complying with demands by a authoritarian regime.
43.
by fisf 2y ago
Because if you factor in the cost of storage there might not be arbitrage.
44.
by fisf 2y ago
Especially since there is no new value generated that could trickle down. Such "rent seeking" is only redistribution towards lend owners.
45.
by fisf 2y ago
> People here seem to think that Nvidia has absolutely no advantage in their microarchitecture design skills. It's all in software or monopoly. That's an extrapolation. Microarchitecture design skills are not theoretical number
46.
by fisf 2y ago
Yes, but this is not where I am getting at. "Will be interesting to see if other high level accelerator supporting languages like Chapel or Futhark or JAX end up getting NPU backends, it might give them a nice boost over the proprietar
47.
by fisf 2y ago
Support in for high level backends almost always comes down to c++ either way. That goes for NPUs (which are really fragmented anyway, i.e. there is no uniform API), and and also for JAX' TPU backend (which iirc is using XLA).
48.
by fisf 2y ago
> reimplementing something poorly from first principles Not-invented-here syndrome in practice. Which kinda works for the n-th shitty web framework. But it goes really poorly in any complex domain.
49.
by fisf 2y ago
Yes, that's possible for product development in a startup (even necessary, because you don't know what you need to build), and in web dev, end user facing apps,... For many other domains, the full incremental / Agile loop is
50.
by fisf 2y ago
> Why not? If you fulfill the role of a supplier of a component in a bigger product (especially in an industrial setting), you cannot "try to build something". People are not happy if you tell them that you want to iterate on t
51.
by fisf 2y ago
> But my fundamental reaction is, what does failure mean. Because in my world view failure should be expected and accepted and learned from. That's fine in some domains. It's not acceptable in others (i.e. you cannot just try t
52.
by fisf 2y ago
Software developers usually have managers of various sorts of supervisors. This is classical division of worker / manager in the Taylorist sense, as in the managers are mostly not developers themselves. This is what makes software d
53.
by fisf 2y ago
> That's part of the addiction too - it can feel meaningful for a long time. That's obviously true of meth,etc. as well. The first step is always to realize you have a problem that's negatively impacting your life.
54.
by fisf 3y ago
Yes but the same could be said about rocm.
55.
by fisf 3y ago
AI startups who just wrap a (standard) API or model in a thin UI layer. The backend part will be a commodity and the UI layer offers no value proposition.
56.
by fisf 3y ago
That's fine and dandy, until you realize that Jax only has a limited amount of backends. E.g. rocm support is still experimental. Somebody has to build those optimized backends -- it's not just a matter of people picking the wrong
57.
by fisf 3y ago
Jim is at tenstorrent now.
58.
by fisf 3y ago
Sadly, some of the countries doubling down on coal, are going to be most severely impacted by climate change.
59.
by fisf 3y ago
> The important thing is what’s being communicated, not the semantic definition. The above comment’s point was pretty clear IMO But the argument is, that Nvidia should follow a standard valuation model, since their product is a commodity
60.
by fisf 3y ago
This. Many people here have a lucrative career in traditional fields, big tech, etc. Working in those fields is good. Building "products" is good (even if that only means optimizing conversion rates and pushing ads). Doing well in

More ›