Before commenting on this it is critical to listen to the whole interview.
http://insidebitcoins.com/news/bill-gates-3-criticisms-of-bitcoin The interviewer does an excellent job of asking the tough questions and Bill Gates attempts but in reality fails to "doge the bullets". The first of course is when the interviewer effectively pins down Bill Gates into supporting GMOs. The real bullet he has to doge is when the interviewer reminded Bill Gates how he benefited from open systems to develop as a coder, and how many systems today are closed. No mention of the dreaded G (GNU) and L (Linux) words but the message is loud and clear. Bill Gates avoids the issue by shifting the focus to basic literacy rather than coding skills. My take is that he knows that many of the answers are to ditch propriety solutions but of course will not admit to that.
Poor people cannot currently rely on bitcoin
“The effort to make sure your bitcoin provider isn’t going to lose your money and your understanding of the volatility of bitcoin — I’d hardly say that it’s ready for, you know, poor people to have it go up and down by a factor of two and, you know — ‘Oops, I was at Mt. Gox. Now that’s not good. Now I’m at Bit-whatever.’”
He is wrong on both counts here. First on the volatility issue there are many situations where the the fees associated with a propriety transaction outweigh the volatility risk even for the homeless.
http://www.wired.com/2013/09/bitcoin-homeless/all/. The issue of the provider risk misses the entire point of Bitcoin. The solution is to not rely on a provider but instead store the XBT on a device that one trusts. This of course brings up the issue of malware, in order to prevent the theft of the XBT. The solution to the malware issue is of course very simple.
Ditch Windows and run GNU/LInux instead. Now does anyone really expect Bill Gates to admit to that!
Lack of transaction reversals
“So that basic technology shows that digital can do these things very cheaply, and the fees that have been building up over time won’t stand up even for small transactions. Now making sure that the thing is fraud-resistant and that money can be refunded – there’s somebody that you call up if you think you transferred to the wrong account or your account balance is not what you’d expect.”
He is wrong again. The problem with this argument is that it only works if the payee has good credit in order to honour the reversal. There is a reason why one of the requirements or of a Merchant account of a PayPal account is a good credit rating. Now if the real objective here is to address poverty a payment system that relies on the payee having good credit makes absolutely no sense since in many cases poor people have no credit or bad credit. After all if it is difficult or impossible to make payments to poor people how on earth can one expect such a payment system to help reduce poverty!
Potential Anonymity
“Also governments, for most transactions, will want attribution, that is, the idea of a system where you can’t see — is that drug money, is it terrorist money? Should that be taxed? You’re going to have some tension between the attributed systems like credit card [or] debit card systems where there’s actually a record of who’s engaging and the purely anonymous ones. The one where I see it getting to critical mass, along with the government regulatory support we need is where it’s attributed; where we can see who actually did this transaction.”
He is wrong again. The issue here is cost vs return pure and simple. The cost of doing AML/KNC on a prospective client is close to the same regardless of the net worth of the client. This net worth can be 0.60 USD or 60,000,000,000.00 USD. The difference in the profit potential for the institution doing the AML/KNC differs by 11 orders of magnitude. For this reason the only digital money transfer systems that actually work and are cost effective for the poor are those that do not have or effectively do not have attribution.