Hacker News Reader

Comment by georgeburdell | Hacker News Reader

As a hardware engineer who admittedly has a bit of contempt for software types, I can't help but think that maybe software startups just aren't novel enough these days? As noted in the article, software has dramatically lowered the barriers to starting a company, so isn't it inevitable that major companies would exploit the same low barriers (powered by their ability to gather intelligence on the next big thing via their platforms)?

View on HN

It's not discussed in this particular article, but economies of scale certainly favor incumbents in hardware as well. If you need confirmation, go ask Pebble guys how they fared after Apple Watch was launched.

I am not sure you can use one example of a hardware company failing as proof of anything. Just as you couldn't use the success of a software or hardware company proving the opposite, that there isn't a problem.

There are enough articles out there from hardware kickstarters that failed post-funding for me to believe that manufacturing new products on-price and on-time is insanely difficult. I'm sure Apple can design and manufacture 50+ iPhone prototypes for any yearly launch. But startups really only get one chance to get it right before their funding runs dry.

You don't even have to look at startups. Look at all the teething issues that Tesla has had ramping up production, versus an established car company. Look at production issues for things from established companies like the Nintendo Switch. Even the iPhone X had its production issues.

Certain segments of software startups aren't novel enough these days. It's pretty hard to start a new consumer web company or mobile app and make any money off it.

Software has the dubiously nice property of its foundations being invalidated every 5 years or so, though. You can still make a lot of money in crypto or AI, for example. Ethereum is full of smart contracts that are doing the same things microcomputers did 40 years ago, millions of times slower, that nevertheless have raised hundreds of millions of dollars for it because they redefined money in a way that somehow convinced people to pay for it. And now all of the nice properties of software like "can process more than 15 transactions/sec" have to be rediscovered, which provides plenty of fodder for entrepreneurs.

Hardware doesn't usually invalidate the laws of physics on a rolling 5-year basis, by contrast. That gives firms an ability to build up large stores of domain knowledge which can't be easily replicated by startups, so it's correspondingly harder to get a new hardware startup going.

Yes, also tall trees in the forest block the sunshine to smaller trees, at times. However, software and hardware need each other and although one may say that hardware is where the true innovation lives, our ability to reason and make practical use of <abstract> data is limited by our understanding of computing science.

why contempt?

I assume it is because he thinks engineering involves making physical objects, rather than abstract ones.

I am a software guy, and I have heard other hardware guys make similar statements as well.

An example is a physical car engine (hardware), vs a video game rendering "engine" (software)

Which some what ignores the fact that a 2 stroke engine has three moving parts in its simplest form.

and mech engineering doesn't have its clients demand 5 months before a new engine delivery you now make it a rotary Wankel engine as that's the new hotness :-)

Ex BTEC A1 (Thermo Fluids ) Eng Tech BTW ;-)

Is BTEC suddenly a notable qualification? And isn't A1 the lowest level (up to 5?). Last thing I heard was that Pearsons BTEC system was a gateway to University in the UK; akin to GCSE or A-Level depending on the level taken. And A1 would be year 10 (14-15 years old) ~ GCSE level grades.. no?

If this is no longer the case my BTEC's are going to get a prominent position on my CV.

A1 was the specialised version of a mech ENG BTECH aimed at the pump, turbine, hydrodynamic and aerospace industries it did cover all 5 levels up to the equivalent of a HND aka first two years of Uni.

All but one of us worked at the bleeding edge in research for the scientific civil service or in related civilian organisations based at Cranfield.

I did do a spec pitch a few years ago to a small engineering company RBR racing - I recon I could saved them a F1 win using the basic skills I learnt at my first job

Sorry, I'm looking everywhere and I can't find any evidence to support your claim.

BTEC A1 is introductory level from all sources I can find.

https://qualifications.pearson.com/en/qualifications/pearson...

https://university.which.co.uk/advice/a-level-choices/what-a...

As I explained A1 is the Version of the course that went from level 2 - 5

Reply on news.ycombinator.com