Artificial Intelligence
Jan 20, 2020

The state of artificial intelligence today

We're living in a time of major upheaval. Technology is radically changing the way businesses operate and serve customers, artificial intelligence (AI) most of all.

Today, AI can process millions of points of disparate data, find unseen patterns and correlations, and deliver powerful, prescriptive insights to augment human decisions. AI can even replicate human thinking – but at great speed and scale – for greater efficiency and productivity.

There are endless possibilities for AI, and adoption has increased over the years. According to the AI Index Report from Stanford University's Institute for Human-Centered Artificial Intelligence (HAI), 58% of large companies say they've adopted AI in at least one function or business unit in 2019 – up from 47% in 2018.

The AI Index Report contains many observations into the current state and progress of AI applications. These insights are useful for anyone who has or is planning to deploy AI in the enterprise.

Below are three highlights from the report, along with some key takeaways.

1. More AI talent is on the way

Enrollment is growing in AI and AI-related subjects, both at traditional universities and in online courses. Graduates from these programs are going on to fill major skillset gaps within organizations. Notably, the AI Index Report shows that the business world has become the largest absorber of AI talent. In 2018, more than 60% of AI PhD graduates went into the business space, compared to only 20% back in 2004.

For those who are hiring, it's important to create an exchange of knowledge between new AI talent and existing employees who have hard-earned industry expertise. This cultivates a “bilingual" talent base that can speak to data, analytics, and AI models and speak to the appropriate business context. Bilinguals are essential for identifying the greatest opportunities to apply AI that will generate real value in the enterprise.

2. Focus on risk mitigation continues

With AI playing a larger role in decisions that impact customers and organizations, there is now increased scrutiny of algorithm and model integrity. The AI Index Report notes fairness, ethical use, data privacy, security, and the ability to explain AI's uses as the main causes for concern. Governments around the world are aware of these risks and have introduced more AI-related legislation as a result.

But what are businesses doing? Well, the report shows that only 19% of large companies say that their organizations are taking steps to explain their algorithms, and a mere 13% are mitigating risks to fairness and discrimination. Ultimately, they must do more.

You need to establish proper governance and ethical frameworks for AI so that you can:

  • Clearly explain how you use data and keep it secure
  • Consider the ethical use of applications and monitor for deviations
  • Trace how algorithms arrive at decisions
  • Control unwanted scenarios, particularly related to fair use of data
  • Only then is it possible to establish trust within your organization, among customers, and with legislative bodies. It's not just a matter of trust in AI, but also in your company as a whole.

3. AI adoption is accelerating

The amount of enterprise data is exploding. Computing and data storage costs are going down. All the while, AI model training and processing is getting faster and smarter. The AI Index Report notes that the time required to train an image classification system has fallen from three hours in October 2017 to just 88 seconds in July 2019. These breakthroughs are shortening the path to AI adoption, helping enterprises create effective and accurate models in less time.

About the author

Vikram Mahidhar

Vikram Mahidhar

Business leader of AI solutions

Follow Vikram Mahidhar on LinkedIn