Alphabet’s Minsky Moment

Marvin Minsky, the American computer scientist, noted in the 60s; "within a generation ... the problem of creating 'artificial intelligence' will substantially be solved". So, it is with great trepidation I write this note. Fortunately, Alphabet (nee Google) has been tinkering in AI since 2012, so I feel somewhat protected from overreach on my predictions.

And I doubt Minsky was contemplating a more immersive maps experience when dreaming about the future of AI in the 60s. However, here we are, and as an investor, I’ll take the new look maps and image-based search over a ‘sentient being’ every day of the week. Throw in sophisticated ad placement capability and you have something that may even pique the interest of Luddites.

And the beauty of Alphabet’s technology is that you don’t even know you’re using AI, whether searching for a new restaurant using an image of your favourite food, using Google translate to convert text or documents or using an AI assistant to get something done. It’s all powered by AI. If you are interested in hearing more, check out Google AI. But if you’re like me, you prefer simple examples and can leave “Sublinear Time Oracles for Hierarchical Clustering” to the boffins at Google.

In this article I explore three examples of AI driving incremental improvements in Google’s service:

1. Visual search
2. Immersive view in Google Maps
3. Ad placement by Performance Max

Visual Search

For over two decades, Alphabet has dedicated itself to its mission: to organise the world’s information and make it universally accessible and useful. The journey began with text search, but over time, it has continued to create more natural and intuitive ways to find information. The latest iteration is Google Lens which replaces the text box with the use of your camera incorporating image-based search. And now with Google’s multisearch feature you can snap a photo of your favourite food, add the text ‘near me’ and Google will show you restaurants nearby selling that dish. The age of visual search is here and the camera is the new keyboard.

And it’s not just handy for finding restaurants. Shoppers can use Lens to select any product and list shops nearby where you can buy it right now. Lens has utility for advertisers and consumers alike and is a key pillar in the future of search. Sellers can list their inventory for free including store hours, directions, phone numbers, product price, product availability and the customer has the certainty of knowing the product is in stock. Google Lens is already answering 8 billion queries every month and is set to significantly exceed these levels.

Immersive View

17 years ago, Google Maps completely redefined what a map can be and today it connects more than 1 billion people every month, and thanks to AI, maps are still being reimagined. ‘Immersive View’ is a new way for users to explore the world, letting them see multi-dimensional views of an area. Immersive View displays critical information like weather, traffic and busyness, as shown in the example below. 

And this is just one example of how AI helps to increase engagement by creating new experiences previously thought unimaginable. As an Alphabet investor, we are reassured by the knowledge that these technologies required tens of billions of dollars of investment, and the competition is a long way behind. If you have an interest in exploring more, check out the recap of Google Search on 22.

AI for Advertisers

Ad placement is the perfect job for AI, as it can process the substantial amount of data generated by YouTube, Display, Search, Discover, Gmail and Maps. At Adweek recently, Alphabet announced product feeds are coming to Discovery, the feed that appears on the home page of Google’s IOS and Android App. Advertisers can use Discovery campaigns to reach up to 3 billion customers across Google feeds to achieve performance goals in Google ads. Google’s unique understanding of customer intent means advertisers can deliver more meaningful ads to people when they’re most interested.

And Google recently released Performance Max, the first iteration of its new goal-based campaigns that allow advertisers to access all their Google Ads inventory from a single campaign. Performance Max creates ads from an advertiser’s assets such as images, videos, headlines, descriptions and logos. The assets are automatically mixed and matched to find the top performing combinations based on which Google Ads channel the ad is appearing on. It’s like acquiring your very own advertising executive to optimise campaigns and automate targeting. Performance Max helps advertisers drive performance based on specific conversion goals, delivering more conversions and value by optimising performance in real-time and across channels using Smart Bidding. 

It is obvious this innovation is a giant leap forward for advertisers. More importantly, these investments are a significant factor contributing to our conviction in the continued growth opportunities for the world’s largest digital publishers. Combine this capability with the tens of billions spent on AI and you have a winning formula for growing revenue over time.

Today Alphabet trades on an EV/EBITDA multiple of 8x FY24 or 10x FY22, well below its 5-year historical EV/EBITDA multiple of 18x. With $116 billion in cash and marketable securities, a strong innovation pipeline and secular growth amplified by the growth in ecommerce, we remain very excited about the years ahead, despite the macro clouds on the horizon.