Will Google succeed on its new Announcements, or Inbox it.

The ultimate season of Game of Thrones is on and fans are already shuddering with apprehension for the next gory death on the show. We are all too used to the show’s writers nonchalantly killing off principal players. But, even before the season premiered, Google seems to have beaten George R.R. Martin to it by announcing the demise of its swooshy AI-powered email management app, Inbox, at Next’19.

Next’19 was a highly anticipated event for all tech geeks, and as far as flair and flamboyance go, Google did not disappoint. It made as many as 29 new announcements about services based on AI and ML and killed off a few of its old offerings.

As the Google graveyard labours under the weight of the piling heap of discarded services, it beggars a look at the crashing of AI dreams even at the wake of new launches. While Google has decided to discontinue some of its other products such as Allo, the tech world seems to be preoccupied with its decision to withdraw Inbox. Some netizens feel that it was definitely not working but many others are quite distraught to see it go. In fact, some publications have even taken to mourning Inbox quite dramatically, with TNW calling it “ritual slaughter” and another tech website dubbing it an “assassination”.

However, regardless of the state of mourning Inbox seems to have inspired upon being cancelled, the death of this service is indicative of a greater reality for me: how the promise of artificial intelligence can fail even in the hands of giants such as Google. In the tech space, artificial intelligence and machine learning are some of the weightiest buzzwords today and corporates are effectively playing to the gallery by stocking up on new AI products. Google is a prime example: having dedicated 29 of its 122 new launches to this emerging field of smart technology. While startup founders may be worried at the rapid and aggressive expansion Google is trying to pull off, there seem to be some cracks underneath the surface.

If you go through what Google calls its “smartest laundry list”, you will find that it has launched a plethora of very specific products and services that use AI, ML or DL to meet precise needs, ranging from tailoring customized recommendations to custom labelling and tagging for videos. However, the question that pops into my mind at Inbox’s unceremonious exit is: do we really need these many products around the same technology or will half of these be killed off again by the time Next’20 comes along? Given Google’s penchant for axing off its own offerings, the latter possibility definitely seems more likely.

Why was Google Inbox discontinued?

The most obvious reason for this decision seems to have been that it is possible to integrate most of Inbox’s trailblazing features into the Gmail App itself, without necessitating a separate email client altogether. As long as it was in use, Inbox was seemingly a testing ground for AI-based features that would eventually be incorporated into Gmail itself. Besides, there seems to have been some allegations about reduced human control as well. As artificial intelligence took charge of labelling and grouping our emails, we had less of a hand in tweaking those labels and bundling emails along different lines.

The latter issue seems to be symptomatic of a persistent problem with AI that even big-shots like Google find hard to solve. As I was recently reading in an article published by the MIT Technology Review (“The Dark Secret At The Heart of AI”), oftentimes these smart technologies can begin to supersede the human influence and make decisions via their artificial neurons in a way that even their developers find hard to understand. While it may be easy to overlook the lack of human flexibility in the Inbox app, the issue can become quite prominent in more serious applications of the AI.

Therefore, while I look forward to seeing how Google’s new AI products perform, I am not exactly delirious with excitement as far as these 29 new launches are concerned. As the demise of Inbox proves, these promising entrants are equally likely to end up in the famed Google graveyard, for not quite matching up to the grandiose claims they were launched with. AI-driven products can easily fall flat even under the aegis of towering names in the tech field and it is best to remain somewhat cautiously credulous of all that they promise.