[lifeformed]

That'd be an interesting advertising alternative that could be good for both parties. In exchange for content, you give the website X CPU cycles (or GPU I guess). It could be built into the browser so that you'd have really efficient mining techniques. Clients (with beefier computers) could choose to give more cycles to their favorite sites.

I'm not sure if you could provide comparable amounts of revenue. I guess time-spent-on-site would be the most important factor. At the current cost and difficulty of Bitcoins, you'd need 1,000 concurrent users on average all day all running at 10mhps to just make $92 per day. 10mhps seems like it's asking a lot, it'd probably drain a lot of batteries.

[crgt]

What if you were, say, Facebook, and had lots of visitors, good time-on-site metrics and you were still searching for ways to grow revenue? Might it make sense? Seems like their entire userbase could combine to mine quite a few bitcoins.. But I'm not a miner, so perhaps I'm wrong? A quick search turned up a Fast Company article that claims (http://www.fastcompany.com/3005269/facebooks-daily-mobile-us...) 618 million daily users. Assuming that most of those folks aren't logged on all day and many (more than half!) are logging on via mobile, you'd probably only be able to convert a fraction of those to effective miners, but even still it sounds like a lot of bitcoins per day. Perhaps they might offer users the option in exchange for better privacy controls and/or an ad-free experience?

[rmc]

Remember that the bitcoin network adjusts itself to try to only have about 25 bitcoin per 10 minutes mined. So if you were to jump into the pool with loads of cpu power you'd get a few bitcoin for a whole, then it'd adjust to be harder, raise the bar, and you're back with the same generation rate

[IanCal]

Yes, but you'd be getting more of the pie. The pie remains the same size, granted, but you can still make quite a lot from owning a significant portion of the processing power.

[kansface]

A ton of companies have tried to monetize spare cpu cycles. The problem is that you can't do anything valuable with them. Most problems aren't trivially scalable, you have to assume malicious users so that means you need to essentially double efforts, and third parties aren't comfortable sending data through the system.