Softbank president Masayoshi Son walked out of President-Elect Trump's office with a big grin on his face this yesterday, which can only signify one thing for me: a Sprint/T-Mobile merger is a done deal.

Actual news stories don't want to say this, because Trump and Son didn't say it explicitly. According to the New York Times, Son "is projected" to spend half of his $100 billion "Vision" investment fund on tech companies in the US, where, the Times notes, he was already expected to spend a lot of money anyway.

You know what else would be a $50 billion investment, though? Buying T-Mobile, which Son has been trying to do for years. His proposed merger between Sprint and T-Mobile in 2014 got shot down by the Obama administration's Department of Justice on antitrust grounds, but Son said as late as this August that he was still hoping to get the deal done. In 2014, he offered $32 billion for T-Mobile, and it's worth more now thanks to John Legere's successful leadership. Wrap in a little bit of capital expenditure building 5G, and you have your $50 billion investment right there.

Don't Trust Masayoshi Son

There are a lot of problems with Son's promises, though, and a lot of reasons why they're not necessarily going to result in a lot of great new American jobs.

Let's take that merger, first of all. The reason the Obama DOJ killed the Sprint/T-Mobile merger in the first place is that it was likely to lead to higher prices for wireless service, as the two big low-cost competitors wouldn't be competing against each other any more. Also, as we saw during the examinations of the failed AT&T/T-Mobile merger in 2011, these kinds of mergers between big, direct competitors tend to lead to massive "efficiencies," which is corporate-speak for huge job losses between departments that are judged redundant.

Son is an amazing salesman, though, and I can see him spinning a tale to the President-Elect that the transformation of the whole industry with 5G will be so disruptive that previous rules don't apply. Son's promise to generate 50,000 jobs, meanwhile, may dovetail with the fact that about 50,000 people work for T-Mobile right now.

If Son invests his $50 billion in other American tech companies, the outcome may not result in a ton of jobs for Trump's job-hungry supporters, either. It really depends on what kind of companies Softbank buys. Softbank spent $32 billion for ARM this year, which has only about 4,000 employees. But Verizon's purchase of Yahoo for $4.8 billion involved 10,000 employees.

Technology companies, and technology manufacturing, also now rely heavily on highly educated workers and automation. As the Times story says, a related (pre-Trump) plan by Foxconn to expand in the US involves a plant to build robots where the jobs would be "higher-end and limited in number." That may not work for the high school-educated folks who formed a key core of Trump's voter base.

So I'd keep a gimlet eye on Son's promises. They may be less than they appear.