Raj Datta, the Global VP for Software and AI Partnerships at IBM, shares the trends he’s seeing in the industry of AI adoption in the Enterprise, and how you can help accelerate AI adoption for small and large Enterprises across various industries.
These are two interesting quotes from larger Enterprise companies, with two themes. They’re experiencing collaboration between companies like never before, and the need to adopt AI early to win.
Why is it so important to think about adopting AI sooner rather than later? With AI, you have the ability to accelerate your revenue 74% faster than your competitors. That kind of incremental growth in revenue really changes the game for companies, and that’s why the race for AI has become critical for every software company.
Key Areas Where Enterprise Needs Help
Everyone’s figuring out how to get started with AI. This is a new area and a new way to innovate, and it’s happening much faster than even Cloud adoption was. As we look at different industry trends, the biggest areas where Enterprise is struggling are:
- Customer service
- The talent lifecycle
An important stat to consider in the customer service space is that 91% of customers who are dissatisfied with a brand will leave it. That’s not good for retention rates, and software companies adopting AI can utilize it to help solve for that problem.
The other area is the talent lifecycle. It’s difficult to scale this space without leveraging AI capabilities. Some other areas where AI adoption is challenging for the Enterprise are:
- Limited AI skill sets
- The high price of AI
- Ethical concerns and not having governance models.
So why don’t we have more companies rapidly adopting AI? Currently, there’s a high price on GPUs out there. Additionally, ethical concerns continue to be a very large area and concern for enterprise companies due to the lack of governance models. We’ve all heard about how AI can spit out biases or incorrect information. That’s a huge concern for enterprise companies because if you’re going to adopt AI without investing the time and money to put in the right guardrails for it, you’ll be constantly very worried about what the AI’s going to spit out. And that is an area that IBM continues to invest in with its customers and within their research teams.
Massive Productivity Gains
What happens when you add AI into Enterprise technologies? Massive productivity gains. IBM sees a 70% estimated productivity gain across itself and its customer.
We’re seeing AI assistants explode, and it’s helping increase employee productivity. At IBM, the majority of their HR is run via AI assistant. You can promote employees, move locations, add salary, and even book travel.
These are things many never thought could happen. What if you transfer an employee via assistant to the wrong place or out of the company? But it works and IBM has been able to gain a 70% operating costs reduction via digital labor. The customer experience is increasing by 10x because they’re happier and love the brand more. And 80% of IT ops and app development time is reduced, which means faster go-to-market.
Models are Becoming More Accurate
45% of the AI models out there are becoming more accurate because they’re open for others to use and train.
IBM has spent a lot of time and effort with research to open up their models. They have the belief that one company is going to use multiple AI providers and multiple sets of AI models –– so they opened it up. And why is that important? When you open it up to a broader community of developers and you have thousands and thousands of individuals contributing to an area and it affects how much accuracy you can get within your models and how much you need to spend on training it just yourself. IBM’s costing has gone down 100% since opening up its models.
Usually, slow AI adoption is the result of costs going through the roof and accuracy. As you build for the Enterprise, key areas to consider are:
- Multimodels. No company will grow with one set of models.
- Data. Everyone knows how disparate data is, and if you can take that data and create the right foundation around it to build off of, you will have the right outputs.
- Governance. We need guardrails within software components to scale our businesses. If the Enterprise doesn’t know where data comes from, they won’t want you in their company. Ensure all AI tech being created is governed so you don’t have biases coming out or incorrect information sent to folks.
The AI marketplace is already here and it’s moving so fast. Companies that adopt AI are growing revenue streams faster than ever before. Instead of jumping in and adopting tomorrow, be proactive in how you do it. Put the right guardrails around it as you expand your business, look at costs, and look to open models.
Key Takeaways
- The AI marketplace is already here. It’s moving very fast and IBM is seeing companies who can adopt growing their revenue streams faster than they’ve ever seen before.
- Especially for enterprise companies, putting the right guardrails around AI is going to be very critical as you expand your businesses.
- Everyone wants AI, but you don’t want to pay the massive GPU cost once your business starts scaling with the AI technology. Open models like IBM’s Granite models can help lower costs as you scale.
