Unlock stock picks and a broker-level newsfeed that powers Wall Street. Upgrade Now
Unpacking AI Agents and The Potential of Today’s Autonomous Applications

In This Article:

The recent and prolific rise of Agentic AI and its offspring, AI agents, represents a poignant shift in how businesses will both operate and innovate. Be clear on this, this is a technological corner we are turning, and as organizations explore and implement, immediate practical benefits are being unlocked. These progressive steps are industry agnostic, and business leaders are keen to find new ways to reimagine their operations.

Unlike Generative AI (GenAI) tools such as ChatGPT, Sora, and Midjourney, which act as on-demand services for various forms of content, from images and video to blogs and essays, AI agents are autonomous. They can understand tasks, make immediate and informed decisions based on logic, and take action to achieve predetermined goals, all without constant human supervision. Now, to really understand how AI agents are being leveraged today, we need to pay close attention to the experiences of companies at the forefront of Agentic AI.

What Are AI Agents Currently Excelling At?

AI Agent adoption is well underway, particularly in cutting-edge technology sectors. Cronos Labs is an early adopter, welcoming the agents’ recent advances across a number of business functions. Ken Timsit, Cronos’ managing director explains "AI agents are already broadly adopted by internal business operations in crypto. For example, at Cronos Labs, we use AI for code generation, code review, invoice processing, summarization of social media feeds, generation of marketing content, and detection of fake users."

It would be a misconception to think that AI agents are only useful for implementation in the cryptocurrency space. Space and Time, working in collaboration with Microsoft, have made progress in other areas of autonomous AI, as Nate Holiday, Co-Founder and CEO of Space and Time shares, "We've been working with Microsoft to create a code generation agent that's already 10Xed our engineers' productivity. It enables our engineers to contribute in a language or area of the tech stack that they're not as familiar with, quickly produce frontend code for easier prototyping/proof of concept, and offload smaller tedious tasks like refactoring to focus on shipping new features."

Holiday added, “Our engineers tell us it’s more effective at generating good, usable, accurate code than any other tools or agents they’ve been able to find on the market, and it’s given us the ability as a company to ship products faster.”

Achieving Balance With Agentic AI

Deploying autonomous AI agents requires the deployer to make many different considerations, one of which is the appropriate level of human oversight. Cronos’ Timsit offers additional insight into this topic, suggesting "The optimal balance between agent autonomy and human supervision depends on the type of task. For creative tasks like creating code and designs, agents can serve as inspiration, and human supervision is of enormous value. For operational tasks, the expected autonomy is much greater.”