source: Flickr/midiman

What? Companies other than Google do AI? This week, and a couple times over the next few months, we are going to look at companies that don’t often make the news about AI, but have invested significant resources in the field. We start alphabetically, with Adobe.

In the spirit of full disclosure

My wife owns a very small amount of Adobe shares, 7 shares to be exact. We aren’t banking our retirement on it and by no means am I using Cloudy to pump the stock so she can make a cool $200. For her to even make $200 as a result of this Cloudy, each member on this distribution list would literally have to buy tens of millions of dollars of Adobe…

If look at the entire 2000s, owning Adobe stock was not a great investment. Adobe was a niche technology player that focused on refining its software suite of creative tools. While creatives swear by their technology, investors were not so enamored. Adobe shares fluctuated in the 20s/30s during most of the 2000s and depending on where you bought the stock during the tech bubble, you probably didn’t make any money at all.

A new era

But in 2009, Adobe did something interesting. They decided to branch out into analytics, buying Omniture, a data analytics company for $900mn. In 2012, Adobe made another decision to switch to becoming a SaaS company instead of focusing on licenses. Since then, Adobe has built a significant SaaS presence in both its creative marketing analytics offerings.

Both moves paid off significantly. Adobe’s stock is up 7x since 2012 and the company almost has a higher market capitalization than IBM. Even crazier, IBM employees 21 times more people than Adobe. TWENTY ONE TIMES. Yes, I say Adobe made some good moves.

Part of the enormous success of Adobe is their investments in AI/ML. Adobe built one of the most sophisticated marketing analytics tools, which include attribution, segmentation and predictive modeling. In fact, when IDC requested marketing cloud vendors to submit AI/ML use cases of their technologies, Adobe submitted more AI-enabled use cases than anyone. In the IDC report, they literally say “there were too many [AI use cases] to fully describe in Adobe’s response to our RFI, so we’ve included a list of additional use cases after the tables.”

But for the all the way Adobe has come, you won’t read any 4,000 word essays or deep dives about how Adobe has transformed itself into an AI company. There is never a profile on the head of Adobe labs on how they are applying cutting edge AI/ML to make their products better.

Why is that?

Adobe doesn’t go on the record and talk about how AI is going to enslave humankind or how its going to revolutionize the way we do everything. The best you can do is Suresh Vittal, vice-president of Adobe Experience Cloud, said AI is going to be bigger than the web. The web is pretty big, but saying “X is going to be bigger than the internet” is a phrase you hear all too often…

Instead of yelling at the top of their lungs, Adobe is using AI in strategic ways to improve its product offerings. Earlier in 2018, Adobe infused its object recognition API to help Photoshop users isolate objects from the background. It’s also used cutting edge generative adversarial networks to produce different makeup effects on people. Adobe is even doing its part to detect fake news by using AI/ML to detect altered photos.

When Adobe makes AI/ML acquisitions, they are also very targeted. In April 2018, Adobe acquired Uru, a computer vision company that supposedly is able to understand what is going on in videos, so that it can embed ads directly in them. A week before acquiring Uru, Adobe acquired a company called Sayspring, which allows users to build and prototype voice apps with no prior experience. Adobe is making these acquisitions to build features of its marketing cloud, so when IDC contacts them in their 2019 vendor review, they can even demonstrate even more use cases.

Bigger ambitions

Ultimately. Adobe is thinking big in the AI space. Two years ago, Adobe introduced Sensei, an extremely ambitious AI powered tool that helps customers with all things digital. Sensei does everything from auto cropping and smart tagging in Adobe Experience Manager to AI-fueled visual stock search in Creative Cloud and Sensei runs across almost all of Adobe products. Adobe even offers an API to Sensei, so developers can build on top of it.

To further develop the Sensei platform, in 2018, Adobe and Nvidia announced a partnership around machine learning and deep learning, focusing on optimizing the use of Nvidia’s GPUs for Sensei. It seems like Adobe has big visions for its AI program as in the press release it said, “The partnership advances Adobe’s strategy to extend the availability of Sensei APIs and to broaden the Sensei ecosystem to a new audience of developers, data scientists and partners.”

Learning from Adobe

What’s impressive about Adobe is they have managed to do this with a relatively small machine learning staff. When looking at LinkedIn data, Adobe has 1,160 people having some “machine learning” function, compared to 12,000 at Google and almost 10,000 at Amazon. Instead of hiring a massive AI force, according to IDC, Adobe sent 1,000 of its developers to receive training on AI so they can contribute in a meaningful way to the Sensei platform.

I think companies that are skittish about AI/ML can learn a lot from studying Adobe. You don’t have to hire an entire university of AI/ML PhDs to start using the technology. AI/ML can be used in small ways to better enhance product offerings or solve real user problems. By starting small, gaining experience, and then scaling up, you can use AI/ML in productive ways.

Now on to the letter B…

Facebook Comments