Artificial intelligence and machine learning are hot terms right now, and many companies are eager to talk about how their products rely on these cutting-edge technologies. Established companies and startups want to explain how AI is a critical part of their business.

But often, “artificial intelligence” and “machine learning” are used as buzzwords, mere fiction used to get attention, client inquiries, and investment from VCs. For this reason, sifting through the AI hype is harder than ever, and business leaders find it difficult to determine who is actually using AI, and who is not.

In this article, we’ll explore three shortcuts or “rules of thumb” for looking at a company from the outside to separate the truth from the hype:

  1. AI expertise in leadership
  2. The use of AI in marketing language
  3. Assessing a company’s investors

This article does not pretend to be an all-inclusive guide to determining who is using AI or not, but we find these approaches to be very accurate proxies in determining true from false AI companies.

Each week we look at dozens of AI companies, and we’re pitched by as many as six so-called “AI companies” per day (on top of all of our consensus research and various in-depth industry coverage), so we have a critical need to discern ream from fake – and I hope that some of our shortcuts will be helpful for you.

(Note: Readers interested in finding artificial intelligence products and services can review our pre-vetted AI companies listings here at TechEmergence.)

Shortcut 1: Look for AI Expertise in Leadership

The first thing I do when I’m trying to figure out if a certain company leverages AI or not – os I visit the LinkedIn profile of their company. I then pull up the profiles of the company’s C-level executives, and figure out if any of them have one of the following two factors:

  1. One is robust academic experience in artificial intelligence or machine learning. What does this mean?
    • Did they go to a reputable university or well-known strong university for a master’s or PhD degree in artificial intelligence or machine learning? Or
    • Do they at least have some like a master’s or PhD in electrical engineering or computer science? These are good proxies sometimes.
  2. The other one is robust artificial intelligence experience at a marquis company. What is a marquis company? A marquis company is a company that is definitely doing AI and leading the progress of AI. These are companies such as Amazon, Facebook, Google, Netflix, and IBM. Do any of these C-level executives have robust AI experience at marquis companies?
Dan Faggella Buckhead Club

Dan Faggella presenting on AI business trends at the Buckhead Club in Atlanta, GA. The “rules of thumb” for determining AI hype were derived from this original presentation.

If no one on the leadership team (most importantly the CEO, the COO, the CIO, or the CTO) has either robust artificial intelligence or data science academic experience, or robust AI experience at a marquis company, then it is very likely the company is just using the term AI and the hype to their advantage, and not actually leveraging AI as a critical part of their product or service.

You might be thinking, “Ok, they say they’re doing AI. Maybe somebody other than the top executives have the qualifications. Maybe the people working under them somewhere have the requisite AI experience.”

That is certainly possible, and you can look for that. However, for a young company to leverage artificial intelligence or machine learning in a powerful way, it usually requires someone on the leadership team to have the AI qualifications and experience.

I’ll explain why this is.

At this point in the technology’s development artificial intelligence is not easy to “do” in business. The science is hard, the math is hard, and the level of data infrastructure and subject-matter expertise needed to derive real value from AI in business are rare. Not everyone can do it, not like building a website or doing email marketing. Hence, if no one in the leadership team is qualified to do AI or machine learning, then in all likelihood they are not deeply baked into the company or the product.

It is not even a question of size or resources. You may find larger, profitable companies with 500 employees that claim to be “doing AI,” and they have one or two senior developers with a bit of machine learning experience.

What might have happened here is the company hired them because they want to use the buzzword. Top executives at large firms often realize the potential of AI (either in practice, or as a buzzword), and thought they should hire some AI people – and BAM – the “about” page now mentions “artificial intelligence” and “machine learning” along with other bluster, like “cutting-edge” or “revolutionary”.

These non-AI leaders probably mean well when they try to “plug in” AI into their existing product suite, but this is absolutely always easier said than done, and anecdotally we believe it’s much more likely to fail than succeed in the near term (i.e. AI won’t actually be leveraged meaningfully in their product, or the process of doing so will take many, many times longer than expected).

It is hard to take a company that has not started with artificial intelligence in its DNA, and implant it after the fact without having any executives that have data science experience. It takes a very long time to revamp an existing product and bring in artificial intelligence.

You are much more likely to have a product or service that really does leverage AI as a core part of its functionality when the executive team has that background versus an existing profitable growing company that hires some machine learning people later to try to do some machine learning to catch up and be “hip.” This does not mean that a startup with AI people in the executive team will necessarily give value. However, it is more likely to get a product that has AI as part of the core product if somebody in the executive team has a deep AI background.

To summarize, the first thing you look for to determine if a company is truly leveraging AI in a meaningful way is robust academic qualifications and/or robust marquis company AI experience at the executive level. If you do not find that, then the company is probably not going to be leveraging AI in a meaningful way, despite what they tout on their homepage.

Shortcut 2: Examine the Marketing and Positioning

ML in Marketing

Our research – such as our recent “Machine Learning Executive Consensus” – involves in-depth analysis of dozens or even hundreds of companies. Often, the “rules of thumb” presented in this article are used as a first filter for sifting out the companies making false claims.

Any business will market their products online, which gives us the information we need for the second rule for assessing an AI company. You need to see how they explain what they sell on the website or the LinkedIn description of the company, and ask the following three questions:

  1. To whom do they sell? (i.e. What kind of buyer benefits from using them?)
  2. How did they benefit that kind of target client? (i.e. What is their deliverable result?)
  3. What is the role of AI in their product? (i.e. Where is it relevant?)

I’ll lay out an example:

A company sells sentiment analysis and text mining software to call centers. The product description on the website states it uses artificial intelligence to comb and mine data for call centers. It transcribes calls, find patterns within the words and within the calls themselves, and determine the kind of sentiment of those calls such as anger or happiness. The software also determines the topics of those calls, such as refunds, delivery problems, buying another product, and so on.

Taking this scenario, let us answer our three questions.

1 – To whom do they sell?

They sell to any company with a call center, such as large businesses with in-house call centers and a reasonably high volume of calls.

2 – How did they benefit that kind of company?

The software will help them learn from the broad patterns of complaints, topics, and sentiments from their customers in real time.

3 – What is the role of AI?

With a large volume of calls, it can be hard to extract data. The AI software takes text transcripts and extracts common topics and sentiments from the actual words in the dialogue and perform an analysis.

It seems relatively clear that the company above has a clear idea of their target market and know the role AI plays to benefit software users. If you go to a company website and you cannot answer these questions, then it is very likely that the company is just riding the hype, and not actually doing AI in any meaningful way. Here are two things you should look out for in a website:

Vague statements

For example, the company website might say “We leverage AI to extract new patterns” but does not specify what those patterns are, or it does not specify how this pattern extraction is meaningful to the user. It might also state “We use cutting edge algorithms to optimize for this and that,” but it does not explain what these algorithms actually do. If you encounter statements like these, then you are very likely looking at a hype company, not a real AI company.

Overuse of buzzword terms

If you read their website and it seems like they are using terms like AI, neural nets, and algorithms too often, you should be wary. They are creating a word salad out of those exciting words, but do not actually tell you what they do. It works for SEO, but it is a bad sign for potential users.

Doing business with suspected “hype AI” companies usually isn’t something I’d advise. Either they have not figured out what they are doing, or they are just fudging their way to get sales or buyer conversations.

Shortcut 3: Examine the Investors

The third rule of thumb in determining a true AI-based company from a fake is to analyze who has invested in that company.

To determine why this step is important, let’s take a step back. If you are trying to figure out if a company is really using AI and their product has any real promise, you will have to do the following things:

  • Meet individually with the executive team members of this company
  • Examine the financials of the company, i.e. income and expense reports
  • Examine the company business model in deep detail over a few months

You probably do not have the ability or time to do any of these if you are reading this article. However, other people do, and those are venture investors. Venture investors have the time to do that because that is their business. They would almost never invest substantial sums of money into a company without a robust due-diligence process.

They routinely do much deeper research into a company than most outsiders can ever hope to do. While some artificial intelligence companies bootstrap their funds from savings and revenue, most require a lot of upfront investment to develop the technology, hire capable developers, and market it. The bulk of AI companies will have venture investment behind them. Most are young teams that have not saved up money for the last 10 years to have millions of dollars in the bank, so they need to raise money. That usually means pitching to venture investors, who will then do the types of research abovementioned.

With that in mind, it is a good sign if an AI company has raised a lot of venture capital. This does not mean that success in raising money through venture investors automatically makes it a real company, or means you no longer have to do your own research into said company. In most cases, however, it is a rule of thumb.

For example, if a company that says AI in the homepage and in LinkedIn, and you see it is backed by big reputable venture investor firms such as Koshla Ventures, Andreessen Horowitz, Atlas Ventures, Charles River Ventures, or Sequoia Capital, you can assume that claims of AI is not a lie.

These reputable venture investor firms do a lot of research on any company in which they invest. If they have given a company millions of dollars, it means they have determined the technology has promise, and believe in the company. You can safely assume that companies like Sequioa Capital would not invest in a company that is predicating their sales on absolute AI hype (as opposed to a genuine and strong AI technology base).

By looking at the type of venture investors for a company, and the amount of the investment, you can determine whether this company is in fact leveraging AI in their sector.

Bare in mind that not all VC-backed firms use AI. I’m not implying that AI is critical for all firms that raise VC funds. Rather, I’m stating that if a firm has raised $20MM from Sequioa Capital – AND that firm claims to use AI – then that claim is much more likely to be true that it would be without the substantial investment from a reputable firm.

Conclusion

To wrap up, here are the three things you can check on a company’s website and LinkedIn page to find out if an AI company has products or services that can actually benefit your business:

  1. Any person on the leadership team or C-level executives of a company with robust academic or robust marquis company AI experience
  2. Clear and concise company/product descriptions specifying the target market, benefits, and role of AI
  3. Reputable venture investors that have put in money in the company

The first rule will actually tell you a lot about the company you are looking at, and you can get some valuable information in just a few minutes. We found after examining thousands and thousands of companies, that is probably the quickest rule of thumb for figuring out if a company is actually leveraging AI or not.

Together, these three rules of thumb can help you determine within half an hour and with more certainty if a company is a bonafide AI provider or not. These are the best shortcuts we’ve found for separating artificial intelligence hype from reality.

I have mentioned on a number of occasions here at Techemergence that only about 20-30% of the companies who claim to do artificial intelligence on their website have some kind of robust experience required to actually be doing AI in the first place. In most cases, it is just a couple of people with MBAs or a couple of Java or C++ developers saying they do AI on their website, but its absolute nonsense.

If you get a pitch from a company who claims, “Oh, yeah, we leverage AI to do this fraud detection stuff,” go look at their website. Go look at their leadership and quickly figure out if these people know anything about AI. Oftentimes, they do not.

However, the rules I discussed are by no means inclusive. I do not claim that you know everything about a company by using these three rules. However, if you are a busy executive with no time to do what we do here at Techemergence, they can help you get a good sense of whether an AI company is real or not without too much effort.

Our analysis and research work in the artificial intelligence space is 100% of our focus here at TechEmergence. It is our business to spend weeks at a time examining a large number of companies in different sectors.

Most business executives that read our content do not have that kind of time. They need quick tips and advice on determining whether a company who says they are using AI is actually using it, and that is what this article provides – and I hope that some of the pointers in this article prove useful as you assess the AI landscape in your own sector.

 

The original transcription and writing for this article (based off of a presentation from Daniel Faggella) was done by  Tini Abadicio, with later editing and additions by Daniel Faggella.

Header image credit: Orbium SA