How to climb the AI Ladder in 4 steps

So what is the AI ladder, and why should my organization climb it anyway?

The AI ladder is a framework developed by IBM to help organizations understand where they are on their journey towards capitalizing on Artificial Intelligence (AI). These guiding principles help organizations transform their business by focusing on four main areas: how they collect, organize, analyze and then ultimately integrate data, machine learning, and AI throughout the organization.

Is your organization ready to embrace the most disruptive technology in the world?

Because if you're not ready to embrace machine learning and AI, you could be the next MySpace. Or Blockbuster. And I hope you understand the importance of AI, because you can't count on your leadership to understand it. According to MIT Sloan a staggering 81% of business leaders do not understand the data and infrastructure required for AI.

Background

Artificial intelligence (AI) allows machines to learn from experience, adjust to new data and perform routine tasks that humans do. Most examples of AI we read about – from self-driving cars to chess or go playing computers – rely heavily on deep learning and natural language processing (NLP). Technologies like these allow computers to be trained for specific tasks by processing large amounts of data and recognizing patterns in the data.

Of course there is a lot of misconceptions around AI. Arnold as the terminator is a part of our collective consciousness. In reality AI already powers many of our day to day interactions with computers and technology. When you need a ride from Uber/Lyft, ask Siri for directions, or waste a weekend churning through Netflix’s recommendations - all of these interactions are driven by computer systems using big data to predict what you need.

Opportunity of a lifetime

Rob Thomas, General Manager, IBM Data and Watson AI, has written that IBM uses the AI Ladder to help organizations understand where they are in relationship to the "one of the greatest opportunities of our time". And Standford professor and Coursera co-founder Andy Ng calls AI the new electricity. "Just as electricity transformed almost everything 100 years ago, today I actually have a hard time thinking of an industry that I don’t think AI will transform in the next several years,” Ng says.

Since the year 2000, 52% of the companies that made up the Fortune 500 have disappeared. They have been absorbed, succumbed to performance atrophy, or declared bankruptcy. Don't be the next one.

To climb the AI ladder, there are four stages your organization needs to execute: Collect, organize, and analyze the data, and then integrate AI throughout the entire organization.

The Basics - Data and Information Architecture (IA)

Have you modernized your data infrastructure?

Today our data is spread across a variety of environments and multiple providers, private and public clouds and good old-fashioned on-premises deployments. To be ready for the business of tomorrow, you'll need to be building a hybrid multi-cloud platform as part of your data architecture. Your organization needs a unified, modern data fabric. AI needs data to be effective, so prepare your data for AI. A data fabric acts as a logical representation of all data assets, on any cloud. Seamless access to all data is fundamental - everywhere from virtualization to the firewall and out to the edge layer.

If you deploy AI projects in any of these environments, you are forced to use the AI tools from those providers. So if you are using Amazon Forecast for time-series forecasting, you are tied to AWS. This is modern day “vendor lock-in,” which can stifle innovation and prevent businesses from scaling AI efforts. Hybrid platforms provide the foundation necessary to support all the capabilities a business would need to build, deploy,and manage AI models at scale.

A challenge that confronts many organizations is that data is distributed across multiple silos, databases, and clouds. But a well built hybrid platform can handle this. Note: If your company is still using Excel and/or Google Docs and Sheets for data or analytics - brush up your resume.

Stage 1 - Collect the data

After your org has modernized it's data architecture, its imperative the data is simple and accessible. Siloed data is of little or no use to an organization. If people and processes can't easily get to it - it's useless. Waiting on a data engineer to transform data so that you can use it? That's a red flag. Be smart and invest in data management platform (DMP). A few more data engineers will not solve complex data problems. You need to use all types of data, structured and unstructured. And you can't effectively train models on only transnational or click steam data. It must be from multiple sources - including 3rd party data.

Now, once you've built this firm data foundation, the company’s IT or data department can give the business reports it can actually use, as opposed to somebody pulling the data out of the system with a reckless SQL query and then trying to figure out how they want to manipulate the data. If your analytics team is spending time writing exclusionary logic to avoid bad data - rather than fixing it at the root level - you might want to talk to your team. Or get a new one. This foundation also sets the groundwork for the company to look ahead at AI and machine learning solutions to continue on its path of digital transformation.

Stage 2 - Organize the data

Data, data everywhere but not a byte to eat. The explosion of data as a result of the cloud and mobile revolutions has accelerated the pace at which we create and record data. Many organizations, however, how no real idea what data they have, where is actually resides, what processes depend on it, and whether or not it is compliant with current data related laws and regulations.

The first thing to consider your data quality. Can the data be used by the organization? Has it been cleaned, normalized, is it complete, compliant, and ready to use to build AI models? When data is not business-ready, finding and putting data to productive use is a constant challenge for everyone - including data scientists, analysts, product, marketing, and finance. If you are still struggling with data quality issues - stop now, and FIX THE DATA! According to Harvard Business Review (HBR), organizations spend 80% their time finding and preparing data for productive use, creating a bottleneck for business agility, competitiveness, and the bottom line.

And you must govern the data. And according to HBR, 70% of employees have access to data they should not, data breaches are common, rogue data sets propagate in silos, and companies’ data technology often isn’t up to the demands put on it. In order to operate in today's more regulated date environment, some of the things that got your start-up here - could land you in jail tomorrow.

Stage 3 - Analyze the data

Once you've been able to collect your data, and have organized it with a trusted and unified framework, you can now use that data to build and scale machine learning and AI models throughout your organization. This will allow you and your company to gain insights from all of your data, no matter where it resides, and engage with data to transform their business—putting themselves at a clear competitive advantage.

In today’s world of regulations, GDPR, and data privacy laws, the way organizations engage with data is under intense scrutiny. Organizations need to manage their data across the entire data life cycle in order to explain either to a consumer or another business how their systems came to a decision and why. This is a serious business. If your engineers tells you that they're not sure where the data is or how it's used - stop, and correct the issues.

Organizations can engage with AI to help them handle the fore‐mentioned issues with predictive insights, real-time analysis, more sophisticated modeling techniques and automation technologies. Again, by investing in a hybrid, multi cloud platform, organizations can accomplish this all in a governed and secure environment.

Stage 4 - Integrate the data

In order for AI to effective within an organization, it needs to be integrated within the entire organization - not just in technology and engineering. That means senior leadership, product, finance and marketing all need people with deep technical and analytical skills. Otherwise, they won't understand the data, or AI - and your company will fall farther and farther behind.

Your customer service team should be using machine learning and NLP to mine your data for it's net promoter score (NPS). Your finance team is engaged in enterprise financial planning. So there is a great opportunity there to use machine learning and AI to use historical and 3rd party data to do predictive analytics. If your organization is still using Excel for financial modeling, you'll continue to fall behind the competition.

Marketing is another area ripe for disruption with AI. This is where AI can actually be at it's best. Finding insights in huge amounts of data that no human analyst could ever find. The most advanced and creative marketing teams partner with data scientists to predict online and shopping behavior, optimizing marketing spend, and delivering content that is engaging, relevant and personal. And with AI, companies have access to their operational data so that they can derive unprecedented insights and improve the efficiency of the entire supply chain.

Conclusion

AI presents an extraordinary opportunity for organizations - and the world. However, no new technology will replace an effective, well-run data-management function. At the end of the day success is about vision, pattern recognition and leadership. Steve Jobs famously didn't wait to ask people what they wanted before embarking on building the next generation of technology and products. He simply went out and did it.

"If I'd asked people what they wanted, they'd have said faster horses". - Henry Ford

Think of some of the most interesting and successful companies today. Google. Apple. SpaceX. Amazon. They all have two things in common: 1) they were founded by extraordinary visionaries and 2) they all have world-class software and data engineering teams that are at the forefront of the AI revolution.

These are the types of companies that you want to work for. Startups and companies run by entrepreneurs or self described geeks. If a lawyer is running your company, that's great - if you're a law firm. But if you're interested in science and AI, want to change the world, look for a company with a tech leader or visionary. Or better yet, a data geek!

At the end of the day, the future is your hands. AI is not only coming to a world near you, it's already here. And it's sitting in a machine right next to you. So be brave, and be bold. Go where mankind has never gone before. Come and join us, climb the AI ladder and see what's at the top.

Joseph Prindle is an entrepreneur, scientist, writer, and fan of the human race. Not necessarily in that order. Please feel free to reach out with questions or comments.