If artificial intelligence confuses you, think about what happens when it’s not leveraged properly. For context, consider the London-based VC firm MMC who found that about 40% of European AI startups don’t use AI in any tangible way. MMC (and TheVerge) are saying that companies just want to take advantage of the AI hype. Meaning that the company or startup “talks” about AI, but they are unwilling or unable to put in the resources to deploy the process, store the data, or make any meaningful use of the information. So begs the question, is artificial intelligence still relevant for startups?
Artificial Intelligence: AI’s promise may be further ahead than its practical reality for young companies and startups that face an uphill grind against their larger peers. Larger entities and new startups have the same needs.
To gather and organize vast amounts of training data needed to build effective AI solutions is cost prohibitive for a startup. Expectations always outpace reality, but that isn’t still a bad thing. There are plenty of entrepreneurs diligently working toward a better future. But in high-tech businesses, it’s crucial that the AI you build, promote, or invest in — is authentic.
Artificial Intelligence: The Substantive Role of AI
The ability to automate responsibilities and streamline efficiency makes AI and machine learning an attractive productivity option for businesses. It shouldn’t be a surprise that the buzz surrounding AI/machine learning and early stage startups peaked in 2018. So much so, that there was a half-joking consensus that adding either to your pitch deck meant an immediate 10% valuation bump with investors.
But the standard for “AI-powered” varies widely — and not only in startups. From a technical perspective, “machine learning” means introducing data into a neural network, so the mathematical model learns to recognize patterns.
Once that AI foundation is in place, the network learns to recognize categorization, transformation, and even prediction.
These capabilities create four startup types, each bringing something different to the AI table:
• Aspirational: Most startups fit into this category, and their founders claim AI/ML deep in the pitch deck. But what those companies mean is that once they’ve found product-market fit and have 500,000 users creating millions of data points, they’ll be able to leverage AI to generate useful insights. None of these startups do any meaningful machine learning work before a Series B funding round.
• Specialized: These startups apply AI solutions to specific industry problems. Examples include Wise Systems, which improves delivery fleets; Standard Cognition, which creates cashier-less stores; and LuminDx, which trains neural nets to identify skin disease better than primary care physicians.
• Foundational: These AI startups build the tools that the AI industry will someday use. Information that will “someday” be used typically means more nuanced API designs or math-heavy algorithmic research. These companies are laying the foundation that the next generation of specialized AI startups will be built upon.
• Opportunistic: AI startups use out-of-the-box machine learning APIs from established tech companies to add a little extra oomph to their products. Identifying whether a cat is in a photo or basic language skills aren’t core to these businesses, but those qualities can distinguish products from the competition. These startups use AI as a standard part of their tool kits, and they represent the future of how most businesses ultimately will use AI.
Understanding which ecosystem a startup fits into is the key to building an authentic AI enterprise.
This pragmatic approach means recognizing when you’re the market leader introducing AI to a new segment and when you’re simply building something that eventually will be AI-ready.
You’re not just adding “AI/ML” to your business to take advantage of a trend. You’re leveraging technology to solve a real problem, which is what makes a business viable.
Artificial Intelligence: When Is AI Relevant to Startups?
When assessing a company’s AI-readiness, it’s important to determine the purpose it will serve. If it’s a support beam for a business, large amounts of data and an understanding of that data’s value to an existing industry are needed.
AI is not a secret sauce — data is the sauce. To that end, it’s crucial to understand how much data good ML and AI requires. The data source is what ultimately drives the ecosystem, and it must be well-structured and optimized. This data also has to be stored securely.
For example, a startup could leverage AI/ML to analyze the entire Twitter firehose t