AI in IT – Cutting Through the Hype
Artificial intelligence (AI) is one of the oldest, most recycled concepts in computer science. Over the years, the goals of AI have remained the same, namely creating systems that mimic human intelligence by:
- Self-configuring and self-optimizing in response to data, i.e., learning.
- Using additional, new data to improve over time and adapt to changing conditions.
- Extrapolate learned behavior to new situations and data sets.
While these objectives haven’t changed, the techniques for achieving AI have evolved significantly over the decades. After futile efforts at cognitive reasoning, logic programming and rule-based expert systems, and a dark age period of little experimentation, AI has been rejuvenated by combining neural simulations with the massive parallel computing power of modern GPUs.