AI, ML, 5G, IoT: Dangerous Investment Acronyms

It's hard to read an earnings call transcript now without running into references to Artificial Intelligence (AI), Machine Learning (ML), 5th Generation Wireless (5G) or IoT (Internet of Things) - even in seemingly non-technical and historically stodgy companies. It doesn't matter what industry you are in: these technologies are going to revolutionize everything, whether you are blowing up enemies, sorting trash or flipping burgers. It also doesn't matter if earnings are bad now, because AI / ML / 5G / IoT are going to provide incredible secular tailwinds for years to come. There hasn't been such excitement and hype in tech since chatbots were going to rule the world.

This chart, which shows venture investment in AI related companies, sums up the trend nicely:

Source: https://www.technologyreview.com/s/612582/data-that-illuminates-the-ai-boom/

Merely mention the words "machine learning" or "data driven decision making" (as if every business decision before this new era was based merely on opinions and emotion), and your valuation will likely sky-rocket. Never mind that most companies can't explain coherently how they will use ML (or even what it is), nor what the actual impacts will be to their bottom lines.

Anyone who has tried to utilize Siri effectively for more than a few minutes will recognize the large disconnect between hype and reality.

The Reality

When most people think of AI, they think of Artificial General Intelligence (AGI) - basically, the type of "human level" artificial intelligence and "consciousness" that is commonly portrayed in pop culture. However:

...none of the scenarios described above - the immortal digital minds or omnipotent superintelligences - are possible based on today's technologies; there remain no known algorithms for AGI or a clear engineering route to get there...Getting to AGI would require a series of foundational scientific breakthroughs in artificial intelligence, a string of advances on the scale of, or greater than, deep learning - AI Superpowers: China, Silicon Valley, and the New World Order

There is no doubt that these new technologies are going to change the world, but these changes are likely going to be more gradual, quotidian and incremental in nature. Similarly, not all industries will be equally affected: tasks that "can be optimized based on data" will be replaced, but AI is still unable to "engage in cross-domain thinking on creative tasks or ones requiring complex strategy, jobs whose inputs and outputs aren't easily quantified":

Source: AI Superpowers: China, Silicon Valley, and the New World Order

The Winners

The big winners of these technological waves are likely to be concentrated in a handful of companies. For example, in AI / ML:

Among the Seven AI Giants [Google, Facebook, Amazon, Microsoft, Baidu, Alibaba and Tencent], Google - more precisely its parent company, Alphabet, which owns DeepMind and its self-driving subsidiary Waymo - stands head and shoulders above the rest...In terms of funding, Google dwarfs even its own government: U.S. federal funding for math and computer science research amounts to less that spending spree has brought Alphabet an outsized share of the world's brightest AI minds. Of the top one-hundred AI researches and engineers, around half are already working for Google.

Many of these companies already offer "AI as a service". Given their unrivaled access to data and talent, they will likely dominate this field. If a company tells you that they are "using AI to revolutionize X" and they are simply piggy-backing on a service from one of the "Seven AI Giants", then how is that a sustainable competitive advantage? And if they say they are developing the technology on their own, even if it is in a specific domain, how likely is it that it will actually be superior to a publicly available, generalizable service offered by the Seven Giants?

If a company is touting their capabilities in AI / ML as a sustainable, competitive advantage - or as the fundamental driver of their business going forward - be wary. It may be true, but the burden of proof should be high. Likewise, if a company is bragging about how 5G / IoT adoption are going to improve their business prospects, the first question to ask is why these same trends won't also help their competitors; advantage is relative. (And it should go without saying to beware the unrelated company that pivots illogically to try to profit on such fads)

Opportunities

I'm by no means an expert nor a prophet, but I have been looking for tangential opportunities in the following areas:

  • Undervalued and under-appreciated companies that supply the "plumbing" for the new infrastructure. For example, Marvell Technology (MRVL) and Nokia (NOK) for data processing, storage, and networking equipment; MU for memory.
  • Companies that have access to a unique, exclusive data set that may enable new, adjacent commercial opportunities. For example, LabCorp (LH) has access to medical data for hundreds of millions of Americans, and is already using this database to help identify promising candidates for clinical trials and drug development.
  • Boring, entrenched traditional companies where the threat of new competition is small, but where even marginal and humble realizations of commodified AI / ML / 5G / IoT automation might materially increase productivity and margins.

Above all, I am trying to avoid companies which do not have fundamentally sound business prospects outside of the "promise" of AI / ML / 5G / IoT / etc. These buzzwords should never be the reason to invest in a company, but merely a bonus if they live up to the hype.

Disclaimer: I am long MRVL, NOK and MU.

Show Comments