Fellow political journalist, and tech executive from San Francisco here. You’re absolutely right, a lot of the focus in Silicon Valley right now is about automation, machine learning, and big data right now (hell, I’m even working on a cloud computing software as a service automation startup right now). The advancements and progress being made in just the past 12-months alone has blown the last 10-years out of the water, and that is just what we are aware of publicly. The impact this is going to have on jobs is going to be insane. I’ve already seen chat bots that can replace most service and sales jobs. Ironically, the labor jobs will be last to go as they require hardware automation which is a bit more expensive to implement. The first jobs to go will actually be in the services sector, which is pretty much been the only source of reliable jobs for Millennial’s since the 2008 market crash. Those jobs will be phased out in the next 1–5 years as everyone moves to self-serve business models, where as manual hard labor work will probably be 5–15 years. As 3D printing becomes more available, we’ll also see a complete loss of manufacturing jobs as well in the next 5–10 years. Meanwhile, it will set the average person back about $16k just to even get a computer powerful enough to seriously experiment with machine learning, and actual AI. Let alone the cost of actually applying it at scale in any useful business context. Granted there are services you can leverage that are cheaper, and you can of course lease computing power online for a fee, but then you’re just paying to use someone else’s computer and don’t actually own it. The most important resource today, and in the future is computing power. The more computing power you have, the more data you can process, and the more data you have to process the more machine learning can learn from it. The reason Google is Google is because Google has the processing power to do what almost no one else can. It is why they essentially have no other competition. The other major owners of computing power are Amazon (AWS), Facebook, Apple, and variety of infrastructure and data center companies in other countries. The consolidation of compute in the hands of just a few companies globally is a major concern, and has already created significant wealth inequality. To actually own AI, you would have to own the hardware itself that powers it. To own the hardware itself the public would have to invest in data centers that rival the major tech companies and I just don’t know how that’s going to be feasible as they’ve spent the last 15-years and billions of dollars building these data centers (Apple’s is even off the grid and is mostly solar powered). Even the US Government pays Amazon $600 million dollar contracts to use AWS because they themselves can’t build enough computing power to compete with Amazon. Honestly, it may already be too late. As a former startup CEO, I can’t even begin to tell you how hard it is to compete with major tech companies that can simply turn your business off with the flip of a switch. In fact, I’ve already seen that switch flipped so many times in the last few years I stopped keeping track. The American people, and the people of the world really have already lost control of the Internet, and it is already being censored by AI (and it was being pretty effectively censored before AI). The only reason sites like Medium even exist is because Ev Williams has money, and a reputation. Most writing jobs in the future will be written by machines, and much of the news media propaganda will likely be determined by trending machine learning data. The US would have to elect a technocrat with New Deal type values to actually pull something like this off, and I just don’t see that happening any time soon. The only thing that gives me hope for the future is cryptocurrencies, and the possible decentralized solutions blockchain can provide. But, even cryptocurrencies and blockchain technology are being invested in by hedge funds and billionaires now, so even crypto might end up in the hands of the powers that be eventually. Although the opportunity to get in now still exists, that opportunity is closing quickly as the US might attempt to regulate. Anyhow, while I agree with your well intentioned article, I fear it is already too late. I mean even Donald Trump used machine learning to help him win the election with Cambridge Analytica, and 2020 is going to be technologically insane. We already live in a cyber punk dystopian reality, the only problem is most people are not tech savvy enough to fully understand what is happening. Let alone be able to vote in any meaningful way to make technological decisions in regards to AI.
