This week I was feeling nostalgic and wanted to see how past predictions fared. Some things I got right, some I got wrong, and some are being proposed in the Senate. Let's start with something I got right.
In Feb 2017, I said the following
While AI/ML used to be reserved for those with sophisticated programming skills, large computing power and massive data sets, AI/ML is moving to a point/click/deploy method where anyone can use it.
Point/Click software, or software where you upload data and a computer builds machine learning models for you, are on a tear. Companies such as DataRobot and Dataiku provide AI/ML point/click solutions and both have raised hundreds of millions of dollars since I wrote that. Another example, Alteryx, which provides point/click big data management solutions has seen their stock increase 6x since 2017.
Although - it hasn't developed the way I thought. I thought large players like Google and Amazon were going to be the leaders in this space, but their products have performed poorly. Amazon's service, SageMaker, only recorded $11mn in sales for 2018 and their image recognition API only had $3mn in sales. According to HG Insights, only 32 companies use Google's automated machine learning solution
Score one for the little guys.
Theory that I am still a believer in, but will take longer that I think: Cloud Everything Theory
Towards the end of 2017 I wrote the following:
I have this theory that everything will move to the cloud. Ok, well, maybe not your dog, but all your digital files will eventually be there. Your programs will mostly run through the cloud. Your computations done in the cloud. All cloud, all the time.
One area that I was bullish on was Chromebooks, Google's low cost computer that stores most of its information in the cloud. I said
there is no need to outfit everyone with $1,500 MacBook pros when you can do the same thing with a cloud connected $200 tablet.
Obviously, we are not there yet and Chromebooks have a barely noticeable market share. But, Chromebooks for education have skyrocketed, with 30 million Chromebooks used for education in 2019, up from 5 million in 2018. My take is as kids grow up with Chromebooks, they are going to shun the 7x more expensive laptops in their later years. We just have to wait 15 years to see if this will play out, but I'm in it for the long haul.
Thing that I was overly optimistic about that won't happen soon - autonomous cars
At the end of 2017, I wrote the following:
Are you including advertising in autonomous cars in your 2020 Tokyo Olympics brief? You certainly could be.
That was a stretch. Japan just recently proposed regulation and rules around self driving cars. Even in the US, where companies are leading the race, there are maybe 100 self driving cars in two marketplaces. There could be automated shuttles at the 2020 games, but you may want to scratch it from your brief.
Issue that I wrote about that could become a law: Value of data
Last year I wrote about the value of data for companies, and tried to answer the specific question: "How much would you pay for Google’s data? " To answer this question, I said
One way to evaluate it would be looking at the amount of intangible assets a company has on its balance sheet. Google currently carries about $19bn on their balance sheet, which obviously seems quite low. Considering Google has almost $60bn in property, plant and equipment (essentially hard assets), Google values storing their data more than the actual data itself. Facebook also carries right around $19bn in intangibles on their balance sheet. Talk about collusion…
Now, we may get an actual answer. This past June, two Senators introduced a bill that specifically would require some companies to "disclose types of data collected as well as regularly provide their users with an assessment of the value of that data." My Cloudy certainly influenced them because the press release said "data increasingly represents one of the most valuable, albeit intangible, assets held by technology firms, "
Finally, the thing I nailed
In the first annual CLOUDYS, my end of year award show, I gave a Cloudy for “The Biggest Bubble of the year Award” to
People writing about how Bitcoin is a bubble
From being referred to as the bubble of all bubbles to arguing about what kind of bubble bitcoin is in, the largest bubble of 2017 is writing about how bitcoin is a bubble.
There isn’t a great way to measure it, but if you compare people searching for “Bitcoin Bubble” to the Actual price of Bitcoin (Indexed to 100), you will see an interesting trend.
Yea, people writing about the Bitcoin Bubble was THE bubble.