Chumby's Chinese Factories (read previous posts too) | Hacker News Reader

Chumby's Chinese Factories (read previous posts too)

bunniestudios.com

57 pointsjoshwa25 comments

Just for a little bit of trivia, Andrew "Bunnie" Huang is the guy who originally cracked X-box while at MIT. The guy is _ridiculously_ smart and I'd recommend reading his blog (it's quite interesting).

Probably the worst legal job available in my town (poor pay, terrible hours, etc etc) is for a factory which makes more cell phone camera lens gaskets than any company in the world. These come off the production line in sheets of a thousand. Your mission, should you choose to accept it: tweezer them off the sheet and onto the waiting chasis. Your quota: a thousand an hour.

There is an acceptable rate for misplacement, which I shouldn't tell you: suffice it to say that my region credits its anal-retentive attention to detail with allowing other, lazier Japanese people to take over worldwide automotive industry.

They have a waiting list for this job a mile long at the moment (my neck of the woods is hurting in the current economic environment).

for those of you, like me, who have no idea what product the original post is talking about, I found the following on boingboing

Chumby is an exciting startup that's making a small, WiFi-enabled, squeezable bean-bag computer that is a delight to hack and play with. You can design your own soft enclosure, install your own OS, write your own apps, or just download programs that do everything from play music and video to run your Chumby as an alarm-clock.

Chumby's co-father is Andrew "Bunnie" Huang, the hardware-hacker virtuoso who broke the crypto on the Xbox and paved the way for all the Xbox hacks that followed (documented in his extraordinary book Hacking the Xbox). Bunnie is a passionate advocate for hardware hackers and software hackers, and it shows in Chumby's open design.

http://boingboing.net/2006/09/11/chumby-chairman-inte.html

Great info, but I thought that I'd supplement that, noting that they also have the Chumby One, which is a hard shell version without included battery. Mine is placed mainly to stare at as I dumbly eat cereal first thing in the morning http://www.chumby.com/pages/compare

And of course the main website is http://www.chumby.com/

I bought a Chumby One for my place and also as xmas gifts for my family. Everyone loves them. It's worth it for the music capabilities alone.

I like how he sums it all up:

"Thanks to him, chumbys are $2 cheaper which frees up more money for us consumers to spend on $2 coffees at Starbucks."

Extra great for me, since I was sipping a $2 Startbucks coffee while reading it...

My prediction that the current savings slowly erode over time as cost of living rises in China. Hopefully, this will allow automation to come into play in production facilities.

Likely at that point, they'll probably just ship basic assembly off to other countries where the cost of living is still lower...

Is no one else a bit skeptical about the quality of the products? Having 1 person eyeball PCB connectors with tweezers doesn't look like good manufacturing and probably prone to shorts.

http://bunniestudios.com/blog/images/mic_skill.jpg

I don't know about everyone else, but I would rather pay the extra $0.95 out of a $119.95 chumby.

I think you can trust that any quality problems are quickly identified long before any products leave the factory. Read up on the MicroSD failure blog post on the same blog for an interesting story.

No, because that one person, having placed thousands of connectors, is most likely better at the job than you or anyone you know. Also, these are connectors - it'd be obvious in testing if they didn't work.

Not everything is reliably tested a couple days after manufacturing. many people have had capacitors gone bad right after warranty, and just because the manufacturer wanted to save $2-3 dollars. http://en.wikipedia.org/wiki/Capacitor_plague

If your suppliers are like this, there is no telling what they will do to cut costs. A cursory test right out of production is not panacea.

Hindsight is 20-20.

and since I write some thousand lines of code every day I don't need anyone to review it.

Also, I'm writting code for a crucial part of the project, so it would be obvious if it don't work. So no silly unit test or anything.

I'd say quality is more determined by the regulatory policies and the accepted standards of the country rather than individual attention/inattention to detail.

Right now as the article explains it's cheaper to use humna labour to do many manufacturing tasks, especially for low volume orders. What I find interesting to think about is the cross over point where it does actually become cheaper to use machines for most or all manufacturing. I think this is all but inevitable. What happens to a country like China when this occurs?

They start making those machines? Or the machines that make the machines? Etc etc.

The same thing that happened to every other industrialized nation: they automate and those people either find other work or starve.

How many typists did word processing software automate out of their jobs in the 80's? Easily tens of thousands just here in the US. What are those typists doing now?

China already produces a lot of people with higher skills (design & development, etc...). It is probably going to be fine.

African countries should start to worry - the time when the development goes from

commodities -> manufacturing -> development

to industrialize your country is coming to an end. In most African countries there are no industrial development (taking advantage of low cost labour). The economy is commodities& subsistence farming.

The short answer is that it hurts. I still remember the moment I managed to grok this, some 10 years ago.

Say some people manage to design machines which can build stuff a whooping 20% cheaper. Because it's an obvious thing to do, you'll have all the factories in the country switch to these machines, and fire a majority of their workforce. Now, on one hand you have a 20% increase in productivity and on the other hand you have 50% unemployment.

You don't have greedy capitalists making money: most of the 20% goes to the consumer, thanks to the competitive market. The biggest share of the money probably goes to the machine builders, but they earned them fair and square: they innovated and thanks to them the job can be done better, easier and cheaper. But overall you still have huge unemployment.

So what is the answer? The answer is a shift in the workforce away from manual-jobs. Simply put, continuous education. It's a fact of life that almost nobody is truly safe in this day and age. Not programmers, nor prostitutes. We could all have our jobs innovated away from us. But this is only a bad thing for those who let the current job become the sole means of paying for the mortgage.

How on earth do you call that skill? He's repeating a small set of steps innumerable times. That's not skill, it's being fast at one thing. We tried treating people like machines in the West a century ago, and over time, it doesn't work worth shit.

I thought this article might have been satire, but apparently not...

That's not skill, it's being fast at one thing.

Being fast at one thing is definitely a skill, if not a talent.

http://www.nbcolympics.com/

I take your point, however, the whole point of this type of narrow work breakdown is to be able to train someone unskilled to do a small, single task well. In other words, this may be a skill, but you wouldn't consider the worker skilled.

One theory as to why China never underwent industrial revolution in the 1700s was the high efficiency of manual labour, which got them stuck in a local maximum.

http://en.wikipedia.org/wiki/High_level_equilibrium_trap

Reply on news.ycombinator.com