Hacker News Reader

Comment by hinkley | Hacker News Reader

That’s not the only one it’s just the easier one.

When people tear down new models of gaming systems they find fewer chips on the boards because they’ve found some chip that does two things for less than the price of both - below a certain point it’s financially infeasible to make simple parts because the capacity they consume makes a chip that does half as much only 15% cheaper. See also when people started putting Linux rescue disks into the EEPROM, because appropriately sized ROMs no longer existed.

You can find better equipment and processes that get you more product per hour out the door without necessarily making the product shittier.

View on HN

Parent comment explicitly says this happens when a sector runs out of things to innovate on. Chips are still innovating like crazy which is why we're seeing some amazing processing/CPU/GPU/DSP chips/etc. If we ever hit on the limits of moores law, watch those CPUs become shittier and shittier over a 10 year period.

I think this point is the key and kind of subtle point - "growth at any cost" does actually rise the tide and bring up all boats when there is a lot of room for innovation. It just starts harming when a sector has diminishing returns on innovation. So you'll be able to come up with plenty of counter examples where growth mindset is really beneficial because it's very context sensitive.

> runs out of things to innovate on.

I suspect you know better than this. Failure of imagination does not mean lack of options. People in our industry reach for next steps that cut off other options in the process, all the time. To assume manufacturing doesn’t have the same problem is questionable.

Sure, but it's a hacker news comment, not a book. To be more clear I mean functionally run out of things. To be even more clear, when ROI on innovation becomes less than ROI on simply making product cheaper. There's always innovations, and some may even be missed, but when the perception of the industry is that future innovations cost more than the loss of good will of making worse product then it's obvious what will be picked.

I think the solution is to make "loss of good will" valued higher, right now it's valued very low by investors since people want to hold investments a short time and exit quickly. CEOs elected by boards controlled by shareholders are the same way. Short term growth is incentivized and that undervalues long term retention of customers, which makes it more likely "manufacture a worse product" a better option than "innovate and make a better product that more people will want".

Yup, the thing is the price difference between the humongous flash (do they even do EEPROM anymore?) and the "appropriate" one is less than the overhead of designing/making/handling different sizes. And the bigger the chip the more sectors for wear leveling and you can tolerate lower total write counts on the chip.

Thus we end up with circuits designed to feed the reinforced concrete outhouse in the center. And they're unrepairable because the troubleshooting cost is more than the replacement cost.

Reply on news.ycombinator.com