How to successfully apply artificial intelligence in a company: What can people learn from machines?

Maja Bilić, a senior product manager at Google Cloud, shared insights on the successful implementation of artificial intelligence (AI) within companies, urging a shift in perspective from viewing AI as a threat to recognizing it as a valuable ally for employees.

Speaking at the DIDS conference, Bilić addressed the common fear that AI will replace human workers. She argued that the true role of AI is to assist humans in performing tasks more efficiently. However, she stressed that for AI to fulfill this purpose, it must be used correctly. "In essence, these are machines that learn from people, and that's why we should use artificial intelligence responsibly," Bilić stated.

“

In essence, these are machines that learn from people, and that's why we should use artificial intelligence responsibly.

Bilić explained that large language models (LLMs), the technology behind many AI tools like chatbots, are inherently non-deterministic, meaning they do not always provide the same answers to the same questions. She pointed out that effective communication, or "prompting," is crucial for obtaining satisfactory results. "The chatbot cannot read the user's mind," she noted.

She used an analogy of ordering coffee to illustrate the importance of detailed prompts. Simply asking for coffee might result in any cup, but specifying the type, cafe, milk, and sugar ensures you get the exact coffee you desire. Similarly, with AI, providing clear instructions, context, and even assigning a role to the AI can lead to much better outcomes. This approach, she believes, will encourage employees to embrace AI as a partner rather than an adversary.

“

The chatbot cannot read the user's mind.