882 karma · joined May 8, 2017
ML practitioners can - by and large - build affordable GPU rigs. That’s why Google is heavily pushing cloud TPUs instead - they significantly outperform GPUs and you can’t buy them (retail or wholesale). You might get some spillover but I can’t see them making an investment as large as Stadia purely to soak up extra ML GPU capacity. If anything, it would be the other way around.
While Storm's Clojure implementation served it well for many years, it was often cited as a barrier for entry to new contributors. Storm's codebase is now more accessible to developers who don't want to learn Clojure in order to contribute.
Very common story that I hear from a lot of shops that tried to go Clojure first. Great language, but too little penetration amongst developers. Makes it very difficult to effectively recruit employees/contributors.
At some point, you just accept that it's more important to get a new warm body than to continue pursuing some idealized programming perfection.
But I’m more interested to know - how does a non-Japanese relocate to Kyoto to work on a startup? Aren’t there visa considerations?
Some markets just don't exist, period. Think nail polish for pet turtles. It's easy to figure this out, because every pet turtle owner tells you "I wouldn't buy this". There's no two ways about it. You know exactly where you stand.
This is infinitely better than the alternative, which is selling a product that kind-of fits, to a market that kind-of exists. You've identified a problem, everyone is saying "great idea, love the design, etc" - but hardly anyone's buying. They all stall, or come up with excuses as to why they can't buy, even though they say the product is great.
Both are cases of "no market demand", but the former is an infinitely better place to be. You know exactly where you stand and what needs to be done - scrap everything and start over.
The latter is a very dangerous place to be. It's easy to give yourself any number of options - "maybe we just need a (UI redesign|new website|better marketing materials)". But fundamentally, the product is flawed. You haven't really found a market need. Odds are you've identified a market want. This is something that's nice for people to have (and that they might be happy to use). But it's not compelling enough to reach into your pocket and pay for.
This is an acute problem in software because
(a) we tend to be builders at heart, so we favour making things over identifying customer needs.
(b) it's relatively cheap to build something. Other industries have significantly higher barriers to entry, meaning you (usually) need to secure customers and funding before the production/manufacturing phase.
There was definitely an explosion of convenience stores across Adelaide - perhaps 2010 until 2015? But they then all shuttered in quick succession.
By contrast, my father came from a much laxer 5 child family, and was the only one who went to university. I can't imagine how anyone would manage with 12.
It’s been at least 15 years since I read Orientalism, but I recall that being exactly my impression of the book.
I came away thinking that it was an interesting collection of perspectives on historical Oriental relations with the West, but that its thesis was simply “You don’t understand us”. I didn’t feel I learned anything to counteract this apparent misconception that we Westerners all hold.
I've ranted many a times about the hypercapitalistic race-to-the-bottom approach that Valve/Epic/etc are taking. This is a breath of fresh air and they have my full support.
But his capacity for logic doesn't seem to be affected. English isn't his first language, but he still writes in full, grammatically correct sentences that are very clear. Wordy and emotional - yes. But not mentally ill.
I think we should be careful not to dismiss/overlook people because we don't share their emotional perspective on the world. In other words, just because he's a unique character doesn't mean his take on set theory should be ignored.
From a very brief click through (and from my own fairly superficial memory of the topic), it doesn't seem to be extremely unconventional. At the very least, it's not standing shoulder to shoulder with Timecube.
But rather than ruminating at length on "PMF", I just prefer saying "if people ain't buying, change what you're selling".
On the other hand, I can see how it's conceptually more relevant to companies taking the route of "raise capital first, get exposure second and figure out profitability later".
https://www.ccn.com/zoom-ipo-stupid-investors-made-unrelated...
I disagree, however, that it was a mischaracterization of, or irrelevant to, the specific comment I was responding to.
I don't necessarily condone Damore's action (or Google's reaction, for that matter) but I'm incredulous that someone could suggest this is not accurate (well, except for the fact that it wasn't an individual woman). I'm interested to hear why you think this is inaccurate.
I also feel it's absurd to characterize it as flamebait. It's directly relevant to the question under discussion - why gender issues are making men feel uncomfortable in the workplace. That doesn't mean their discomfort is justified - maybe you agree, maybe you don't. But flagging these comments indicates you feel that noone is even *allowed to discuss it openly.
Doesn’t matter whether you agree or disagree with the firing, but that was unquestionably the reason.
Unfortunately I only read it after making all the mistakes it lists - but I know from personal experience that everything it says is correct.
When I'm the client, I feel fixed-fee gives the contractor an easy backstop/fallback position - they figure they can half-ass the job and use the additional time to scout for more work.
When I'm the contractor, it's the same, but in reverse - fixed-fee makes clients feel like they can demand the moon beyond spec at no additional cost.
Hourly/daily rates make sure that BOTH parties have something at risk, which tends to be keep everybody in line. The contractor's at risk of the client firing them. The client's money is at risk if they keep demanding endless revisions.
Actually a good lesson for business in general.
As someone with little experience in hardware, what kind of cost would you be looking at for the oscilloscope?
The easiest way to retrain yourself is to stick to simple rules about when to eat (and what to eat).
Also interesting, on an unrelated note:
> Romero is now gone from id. > There will be no more grandiose statements about our future projects.
All this means is that people take all the customer data before you submit your notice.
Different story if someone's fired for cause, of course.