The next leap in AI is arriving – are we ready?

0
83

Arthur Goldstuck | CEO | World Wide Worx  | Editor-in-chief | Gadget.co.za | mail me |


Will machines overtake humans in 2025? Cisco has an intriguing forecast for Artificial General Intelligence (AGI).

The era of AGI, regarded as the next evolutionary leap in AI, is on the horizon. AGI represents a form of AI that can perform any intellectual task a human can. This marks a significant leap from “narrow AI”, which is designed for specific tasks. Subsequently, it embodies fears many have expressed that AI will fully replace humans.

The next frontier in AI

A picture of unprecedented technological advancements has been painted at a recent Cisco online media briefing. In addition, societal shifts and a reimagining of security paradigms were described. These changes aim to ensure AI’s rapid adoption doesn’t outpace its safe deployment.

There will be two types of companies. Those that are AI-forward and know how to leverage it effectively. And those that don’t, and will become irrelevant.

– Jeetu Patel, Executive Vice President & General Manager of Security and Collaboration at Cisco

The above observations cut to the core of the AI-driven transformation. Success hinges on readiness, not just adoption. The optimism about AGI’s potential was tempered by recognition of its challenges. In particular, the concerns regarding safety and security were highlighted.

The briefing also previewed a new AI Defense platform to address the unpredictable nature of AI models. The platform mitigates AI large language model risks, such as toxicity, “hallucinations” and vulnerability to cyber criminals. Indeed, when a model breaks, bad things happen.

The briefing pointed to the inconsistency in how safety measures are implemented by different model providers. Therefore, we advocate for a universal approach to ensure enterprises can innovate without sacrificing security.

The dawn of general intelligence

The AGI breakthrough in 2025 would be as much about technological acceptance as the potential breakthroughs themselves. In fact, humans quickly normalise groundbreaking advancements.

For example, the first time someone uses a Waymo self-driving car, it feels extraordinary. By the third time, they’re complaining about the seats. Therefore, the same will happen with AGI. What seems astonishing today will soon become part of everyday life. More so, the pace of AGI adoption is likened to the internet’s transformative impact.

The first encounter with AGI might elicit awe. But as its integration deepens, expectations will rise. Therefore, Innovation must be matched by the hard work of practical application.

– Frank Dixon, Group Vice President, Security and Trust at International Data Corporation

The future phase of AI

Security would play a key role in unlocking AGI’s full potential. Fears surrounding safety and security would be primary barriers to adoption. Cisco’s AI Defense aimed to address these challenges through a platform-based approach. This approach integrates visibility, validation and runtime enforcement.

It is underpinned by Cisco’s $400-million acquisition of Robust Intelligence. Robust Intelligence is a company specialising in AI model validation and security. It uses a “Tree of Attacks with Pruning” methodology. This method automates the management of vulnerabilities in AI models.

Validation that once took seven to ten weeks can now be achieved in 30 seconds. Instead of bolting on security after the fact, Cisco is embedding it into the development process. This shifts security from an obstacle to an enabler. As a result, developers can innovate fearlessly.

Every application will eventually become an AI application. This shift makes safety and security integral to the AI landscape. Such a shift demands a new mindset. As such, security must be viewed as a foundation for innovation rather than a hindrance.

In conclusion

Just as the software-as-a-service boom spurred growth in the security industry. The next leap in AI will drive demand for robust safety frameworks.

The goal is to ensure organisations can adopt AI fearlessly and focus on the possibilities it unlocks. Behind this push lies a fascinating strategy. Cisco is positioning itself as a leader in the race toward managing AGI while it can still be controlled.

As Patel put it, “We’re moving fast, and we need to move faster.”


The next leap in AI




LEAVE A REPLY

Please enter your comment!
Please enter your name here