Bottom line: Baidu's sale of its mobile game unit represents a broader shedding of non-search assets as it moves into artificial intelligence, though it's far from clear how AI will provide a future business model.

Search giant Baidu (Nasdaq: BIDU) is in yet another headline today that reflects its latest attempt at transformation from its original search business to an artificial intelligence (AI) specialist. This time the development is relatively incremental, with word that Baidu has formally sold off its inconsequential mobile games division for an equally inconsequential sum of 1.2 billion yuan ($174 million).

This particular news comes just days after Baidu founder Robin Li issued a letter to all employees talking about the first official change to his company's mission statement in its 17 year history. (English article) That move seemed a bit overly dramatic to me, and resembled Li's similar talk about putting all his energy into mobile search a few years ago. It all seems to be part of Li's broader personality, that leaves him itching to do something new every 2 or 3 years.

Still, it's hard to teach an old dog new tricks, even though Li is hardly that old just yet. But in terms of China's Internet, he really is getting up there, as are the nation's other first-generation Internet titans like Jack Ma, Pony Ma and Zhou Hongyi. In Li's case, I really do think he's a bit of a one-trick pony, and have serious doubts about his ability to realize his new AI dreams.

Regardless of whether he succeeds, Li is certainly putting into motion a machine that is seeing him rapidly shed all of his company's non-search assets. He spun off his money-losing Qunar online travel business more than a year ago, and tried to do the same last year for his money-losing iQiyi online video business. He's also in the process of reforming his money-losing Nuomi group buying business, which could result in an ultimate sale.

Against that backdrop, this spin-off via a sale of his mobile gaming business hardly comes as a surprise. Chinese media are saying the move was announced in a regulatory filing, which was actually made at the end of March, saying the unit was sold for 1.2 billion yuan to two unnamed parties. (Chinese article) The company will be rebranded as Duoku Games, and will be run separately, the reports say.

The reports point out that Baidu shelled out $1.9 billion for a mobile game company called 91 Wireless in 2013, which was the biggest purchase of its kind at that time. The inference is obvious, namely that the company now being spun off is that same 91 Wireless unit under a different name, and that the unit has lost most of its value in its 4 years under Baidu's ownership.

Bad Track Record

If that's the case, it's certainly not surprising. Baidu has a horrendous record at moving into new business areas, including social networking, e-commerce, music and just about anything else. The Qunar travel service was doing equally poorly when Baidu spun it off in 2015, not for lack of market share but instead for lack of profits. The iQiyi video service is faring slightly better but is still losing money, and Nuomi and the company's take-out dining service look equally miserable.

I do expect that this game unit spin-off which occurred in January, shows that Baidu is serious about shedding all of its non-search units. That means we're likely to see a sale of its iQiyi and take-out dining services by the end of this year, and also possibly a sale of Nuomi. That will return Baidu to its search roots, and give it a relatively clean slate to pursue it new AI dreams.

The only issue, at least as far as I can tell, is that AI isn't really a business model per se, but rather a broader concept that can be included in all sorts of potential applications. Baidu has articulated just one of those so far, namely self-driving smart cars, and is generally considered the nation's leader in that area. But that field is still very young, and I'm far from certain that Baidu will ultimately become a leader in the space.

There are also some AI applications for its core search business, though we've yet to see any of those emerge either. At the end of the day, I would agree with Baidu's plans to jettison its money-losing non-core units that are showing no signs of becoming major contributors to its bottom line anytime soon. But I'm far less certain that its new mission as an AI company has the kind of long-term potential to take the company past its original search roots and into its next stage of development.