Urgent.News

What's breaking now, across thousands of outlets.

AI

The smartest AI might be the one that leaves us alone

One of the ideas I’m always trying to tell my students is that their brains are something, an asset, if you like, they need to protect. The human brain is an amazing optimizer: Provide it with an effortless way to achieve something, and it will quickly start to use it. We used to memorize a ton of phone numbers a few decades ago. Now we can hardly memorize our own, simply because our brain knows…

The smartest AI might be the one that leaves us alone

In the world of artificial intelligence, a guiding principle worth pondering is that the smartest AI might be the one that remains unobtrusive. The human brain, an incredible optimizer, thrives when given effortless means to achieve tasks, yet over-reliance on technology, such as relying on AI to write articles, can hinder personal skill development.

Writing, as an exercise for a professor of innovation, serves to solidify concepts that will later be taught and discussed; using AI to write can lead to a loss of this valuable learning experience.

While AI has its place in tasks such as translation, research, and assessing the clarity of explanations, the ability to write remains a skill that should be guarded. A manager's expertise should not be entirely ceded to an AI that makes all decisions, as this would diminish the judgment and growth of the individual. The key lies in discerning when to trust the AI and when to rely on human judgment, as the latter often involves nuanced understanding and the ability to adapt to unique situations.

When applied to corporate use, the question arises: can an AI learn when its intervention would be counterproductive? A sales agent, for example, might automatically generate a personalized email, discount offer, and follow-up for a customer nearing renewal. However, this may not always be the best course of action, as the customer could have been content with no intervention at all, potentially alienating them and wasting resources.

The challenge lies in discerning when to step in and when to hold back, a skill that can be honed over time through experience and learning from previous interactions.

In the case of AI coding assistants, researchers have found that optimizing for the acceptance of suggestions can lead to lower-quality assistance. Similarly, in the realm of sales, a system that determines when to intervene and when to remain silent is crucial. By learning from past cases, an AI could determine which customers benefit from assistance and which are better left to their own devices, tailoring its approach to individual preferences and needs.

Ultimately, the goal should be to optimize decisions through accumulated experience, rather than blindly executing every suggestion generated by the AI. Proper human oversight, in the form of "human-in-the-loop," is essential to establish obligations and limits within which the system can learn and improve decisions.

Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at fastcompany.com →

More in AI

More from Wednesday 18 November →