Photo Source: CNBC
So Apple made a bit of a splash out of nowhere this week. Google originally released a statement that John Giannandrea, their former Chief of Search and AI, would be taking a step back, but still working on some projects for them. However, it wasn’t long after when Apple turned the tables and announced that they had actually hired Mr Giannandrea to head up their machine learning and AI efforts. It’s a high profile and MUCH needed win for a company that has faced continuous challenges in this space.
The here and now
So what does this one move mean for Apple? It’s hard to say right now, but it could be a game-changer in the long run. Mr Giannandrea rose up the ranks quickly at Google and he was a driving force behind the proliferation of AI throughout all of their products. It’s pretty telling when you and your team end up in the office next to the head of a company.
Still, we have to remember that Mr Giannandrea arrived at Google in 2010 after an acquisition, and it was 2014 and beyond before the public really started to see the fruits of their labor throughout Google’s products and services. While Google’s pace of AI deployment is quite rapid today, it still took time to get there. Bear in mind just how committed they are to using AI and machine learning, and you get an idea of how difficult it is to gain traction, despite the great money and resources.
So how does this translate to Apple in mid 2018? Again, there is no way to know yet. Apple has a habit of integrating acquired hardware technology into their products within one to two years, and even managed to hold to that schedule when rolling out Siri with the iPhone 4S. However, we also know how that turned out in the long run. The task of getting the Siri we live with in 2018 back on track is an entirely different thing.
The fact is this- Apple has already spent an enormous amount of time and money over the last year and a half trying to play catch-up in AI and machine learning. It is also no secret that they have been on a hiring binge lately that goes beyond just Mr Giannandrea. There have also been several company and product acquisitions, responsibility over Siri was shifted to from Eddy Cue to Craig Federighi, and Apple even made some major concessions to their culture of secrecy to aid them in attracting qualified talent.
There have been some short-term improvements in Apple’s overall AI efforts in areas of iOS, but Siri still lags far behind the competition. The lack of progress before the release of the HomePod is frankly a little concerning. However, the current situation just goes to show that this is a long game, and even if Apple is making the right hires and putting the right plan in place to get Siri, machine learning, and AI on track, it will still take time. I wouldn’t expect more than bits and pieces of progress until WWDC 2019.
The long game
While Apple bringing in well known and respected Carnegie Mellon professor Russ Salakhutdinov was seen as a big move and positive sign for Apple in 2016, public-facing progress in machine learning and AI has remained slow so far. The key for Apple has to be coming up with and implementing a plan that will allow them to move forward quickly once they gain forward momentum. Since Siri is the most visible consumer-facing aspect of AI and machine learning for Apple, it has to be assigned a very high priority in this plan.
Maybe such a plan is already in place and we just haven’t seen the results yet. That is certainly possible. Whether Apple is that far along or not, the reason that hiring John Giannandrea is so important for Apple’s future in this space is that he’s actually done this before. He isn’t just a well-known researcher in AI and machine learning. He has actually successfully lead a team that had a significant impact in this space. The right plan is vital, but you also have to have to right person to implement it. Maybe Apple finally got the one they needed. Let’s home so.
A major problem that Apple still faces beyond coming up with the right plan and finding the right person to put it into action is that the competition won’t be standing still while they right the ship. It is going to take a LONG time for Apple to get to a point where it is anywhere close to its primary competition, especially Google. This is why Apple being committed to the long game in AI and machine learning is so vitally important. They will have to prove their commitment to be able to attract talent, and they will have to be willing to greatly increase the time, money, and resources that they are currently investing in the space to start to close the gap. The long game is the only game they can effectively play right now, and they need to play it near-perfect at this point to hope to gain ground on their entrenched competition.
Competence before excellence
Apple playing the long game alone doesn’t sound all that attractive to fans like myself using their devices today. What about right now, or at least the near future? Despite having to focus on the long game, Apple still needs to set some interim goals that make their products at least effective in the shorter term. I look at it this way. Apple needs to make Siri as good as Apple Maps as soon as possible.
Now, if you had told me I would be saying that Siri was behind Maps back in 2012, I would have laughed in your face. Back then, it would have seemed absolutely ludicrous that Maps would ever be viewed as a better product than Siri. However, just look at the evidence today. While Apple Maps still isn’t as thorough or feature-packed as Google Maps, it IS now a very good navigation platform in its own right. Apple invested a lot of time and money over the last six years, and they have turned Maps into a product that looks good, is easy to use, and actually works.
Apple Maps moved from below average to poor at launch to mediocre a couple of years later, to competent a couple of years after that. With its more recent cosmetic redesign and transit and indoor mapping additions, I would say that Maps is actually good enough to be considered above average, at this point. That is a LOT of ground made up over the course of just six years. Unfortunately for Apple, their primary competition just happens to be somewhere between great and exceptional.
Despite that, making Maps into a useful platform that dovetails closely with all of their hardware and services means that more of Apple device owners will use it rather than turning to a separate, more disconnected app from Google. While Apple must to continue to improve it, Maps’ competence at least insures that most of Apple’s users won’t turn away for their mapping needs.
This is EXACTLY where Apple needs to get Siri to within the next two years. The only way for them to get their users to turn away from competing platforms in the short term is to provide them something that WORKS. Siri may be adequate for handling basics on the iPhone, iPad, and Watch, but that is rapidly becoming inadequate. It is already inadequate for a connected speaker like the HomePod.
Apple’s success at keeping customers using Apple Maps shows that many will stay with a fully-integrated service as the default if works at an adequate or better level. Because of this, getting Siri to this level of competency has to be the highest short term priority for Apple. I think even the most ardent Apple fans can agree that it isn’t there today, especially when it comes to consistency. Job one for John Giannandrea in his freshly minted leadership role at Apple needs to be getting this task accomplished, on the way to far more ambitious goals as part of playing the long game against Google, Microsoft, Amazon, and others in AI and machine learning.