The A Word — How We Tell Whether Companies are Using Real AI (Part II)
The startup world is obsessed with AI. Companies are tossing around the term in catchphrases and marketing messages. So many are pitching themselves as an AI company, often without a genuine AI story to back it all up…so how do we tell whether it’s real AI or fluff?
Buzzing About
It’s human nature to latch onto the next big thing: Blockchain, SaaS, IoT, Web 2.0, big data…the tech buzzwords of the startup world can be overwhelming, and largely overused (/misused).
For a while, the startup world has been buzzing around with the A word — artificial intelligence. Barely a day goes by when we don’t come across a company that is pitching itself as an AI company.
Of course, we cannot slash everyone out there that claim they are building an AI company. There are some serious players with deep technical knowledge in fields like data science, that are putting their blood and sweat into developing real AI systems. However, the number of companies using AI for branding purposes is out of hand.
What “Real AI” Is NOT
The AI field has been transforming in supersonic speed, with new discoveries and ideas being developed every day, which makes nailing down a strict definition of the term tricky. (Catch up on Part I of this series to take a deeper dive.) Sometimes, it is better to explain real AI by saying what it is not.
AI is not anything that has to do with data or workflow. People tend to liberally throw around the word “algorithm,” a term that is often associated and have almost become synonymous with AI. But just because a system has algorithms that prompt certain outcomes don’t necessarily mean it is real AI.
In actuality, a lot of these “AI companies” are performing basic data analysis. Their “AI technology” filters data and generates results that are used to drive certain set outcomes.
Many SaaS companies brand themselves as AI companies, but all they do is perform basic data analytics to manage applications and workflows. This technology does not learn from previously analyzed data and does not get more intelligent over time. The technology and functionality provided by such companies can still be hugely valuable and contextually relevant – just that it cannot be considered real AI.
What IS “Real AI”
Real AI is iterative — they are able to “learn” from that information and data they are fed with, and get smarter, increasingly autonomous and capable with more data analyzed.
Real AI is capable of real-time autonomous decision-making, i.e. without human intervention. Many solutions branded as AI are just simple regressions that require extensive human oversight.
What We Look For
Here’s a list of questions we ask ourselves when sifting through the batch of startups who pitch themselves as AI companies:
- Are they just doing more than basic data analysis?
- Are they creating their own data exhaust ?
- Does their system “learn” get more intelligent with more data analyzed?
- Does their technology get increasingly autonomous and ultimately reduce the need for humans in the loop?
If the answers to the above questions are affirmative, we turn our eyes on the founders:
- Technical Knowledge: Do the founders have a solid technical knowledge of machine learning models? Do they know the complex technical details of what it takes to build an autonomous system?
- Education Background: Do they have an education in computer science, data science, computer engineering, machine learning, or related fields?
- Competitive edge: Do the founders have an advantage over competitors? Are they capable of attracting the right talent?
Real AI can provide groundbreaking solutions to real-world problems. The companies that can truly deliver real AI deserve all of the hype coming their way.
(Check out 2 of our portfolio companies — Zectr and Cloudbreakr who have a star team, creating real AI solutions with Marketing Technology.)
A little about Catalyst Ventures: We are a venture capital firm based in Hong Kong. We are always looking to invest in early-stage startups in Travel Tech and Marketing Tech. Ping us if you think we should connect! We love making new friends ;)
