[jonnathanson]

[Warning: semi-tangential deep dive on your point about Amazon not being a monopoly]

Amazon isn't in the retail business. Amazon isn't in the cloud computing business. Amazon isn't in the logistics business. Amazon is in the business business. It is no longer The Everything Store; it is now the Everything Everything. It wants to be the platform around which all of the world's businesses depend.

This is about as ambitious a mission as a company has ever launched, in my opinion -- and Amazon may be the first company with a justifiable claim to such ambition. Its only business constraints at this point are geopolitical, really. I believe it aims even higher in the long run: it is aiming to become the macroeconomic backbone of at least the Western world.

When viewed in that context, traditional definitions of monopoly -- especially the most widely known definition of the state, which is based on market share within a specific industry -- almost feel antiquated. Jeff Bezos isn't JP Morgan; he's freaking Cohaagen from Total Recall.

It is not in his interest, in fact it is starkly antithetical to his interest, to monopolize any single industry. He wants to engineer a state wherein stronger competition within any given market makes Amazon better off, not worse off, because it is the seller to that market moreso than a competitor in that market. (Though it holds de facto monopolist power in at least the publishing business, and probably others, these positions were the stick that makes Amazon's pivot to being in the carrot business credible and effective.)

Concern about Amazon as potentially Too Big to Fail may be warranted, but not yet. I'd argue that the sudden failure of Amazon tomorrow would have a severe but ultimately noncatastrophic impact on the US economy, with competitors like Walmart and Google and Apple ready to absorb the fallout with net-minimal total economic loss.

It is not TBTF yet, though at some point it seems to be on trajectory to get there. Now the tricky part is in gauging what metric we should even be using to talk about Amazon in the next 5 years.

(To be very clear, I say all of this in admiration of Jeff Bezos, not in fear or criticism of him. His business acumen is mindblowing. Whether you think of him as Business Yoda or Business Palpatine is your choice to make. But either way, you have to admit that the force is strong with him)

[sacheendra]

That kind of sounds like Berkshire Hathaway. Or an private equity fund for that matter.

A business which owns businesses.

Amazon is taking the concept to the extreme in that it starts its businesses.

[jonnathanson]

Right, but Amazon doesn't want to own businesses so much as own the business infrastructure.

[petra]

I don't think this is unique to Amazon. Because that transaction costs have become so low(and becoming lower), almost everything is out-sourceable, and naturally in many fields there are advantages to scale, and network effects - common characteristics of infrastructure - a lot of business tasks are becoming infrastructure-like.

Amazon maybe saw it before others. But also, they are in a position that enables them to gradually take more parts of that emerging infrastructure, by owning the customer.

But let's say every stable, low-risk, repeatable task is outsourced outside of businesses into a few large corps. Can new businesses live in that environment? And how will competition look like ?

[bllguo]

in today's culture of the middlemen, Amazon seeks to be the ultimate middleman

[nickbauman]

Last year, online retail sales amounted to 8.6% of all retail sales.