Shuman Ghosemajumder | How your company can get the most value from Al-whether it's Al-focused or not.

Since ChatGPT launched over two years ago, people have fixated on artificial intelligence. To better understand all the hype, it helps to take a step back and tackle some misunderstandings about AI—as well as some history. Artificial intelligence as a field, since beginning in the 1950s, has already achieved a great deal. AI systems have driven advances in computer vision, fraud detection, online advertising, and many other disciplines, primarily through machine learning on large data sets.

- 

Shuman Ghosemajumder is the co-founder and CEO of Reken, an AI cybersecurity startup based in San Francisco, and the former CTO of Shape Security. He worked at Google from 2003 to 2010, where he helped launch Gmail and founded and led the company's Trust & Safety product group.

But what has ignited unprecedented interest in the past few years is a specific type of artificial intelligence: generative AI. GenAI is fundamentally different from the AI forms used in prior decades. Unlike supervised machine learning systems that are used to analyze and act upon new data, genAI trains models on previous data sets to create new content like text, images, and video.

GenAI is powerful and, especially in the case of text generated by large language models, can give the appearance of being all-knowing. And there’s the rub: The AI system may be doing little or no true reasoning, but it offers answers that are articulate and confident, even if those answers are wrong or nonsensical.

Because of this often false appearance of accuracy—as well as continued improvements in actual accuracy—business leaders, especially those without AI or computer science backgrounds, are eager to use AI in nearly every aspect of their companies. Indeed, Goldman Sachs expects about $1 trillion in AI investment over the next few years. But Goldman also has doubts about whether that will drive production of commensurate value. Needless to say, AI must be deployed and monitored carefully to achieve real and measurable benefit.

All of this leads to more pressing conundrums: figuring out how AI affects you and your business, and what opportunities and threats it creates. Here's a handy roadmap as you move forward.