Gregory Dodge – July 5, 2023
The public release of ChatGPT in November 2022 sparked a flurry of diverse reactions.
For many, this stirred up visions of hyper-intelligent AI with dystopian connotations; echoes of “I’m sorry, Dave, I’m afraid I can’t do that” reverberating. Yet, this characterization veers significantly from the current state and near-future potential of AI. To facilitate understanding, we’ll dissect the realm of AI into three primary categories: Narrow AI (ANI), General AI (AGI), and Super AI (ASI).
ANI operates within a defined scope, reacting and responding to specific trained scenarios.
For instance, Netflix’s recommendation system and chess-playing computer programs encapsulate reactive machine learning algorithms, a form of ANI. ANI with limited memory, such as self-driving cars, uses pre-programmed knowledge and sensor inputs to make real-time decisions, refining its responses over time.
The utility of ANI extends beyond these examples, with potential applications in:
1. Real-time detection and prevention of manipulation behaviors, safeguarding individuals from scams.
2. Rapid, large-scale data analysis to identify patterns and anomalies.
3. Assisting subject matter experts (SMEs) in content creation, maintaining their unique voice and style.
Visualize a world where an elderly person clicks a phishing email, but before any harm is done, an ANI in their computer intercepts, recognizing the fraudulent website and blocks the subsequent call from a scammer. The data collected in this process would then be added to an AI knowledge base, mitigating the threat for others.
In the era of zettabytes – digital units of measurement equivalent to a sextillion bytes or a trillion gigabytes – we face an unprecedented influx of data.
By 2025, we’re projected to have over 175 zettabytes of data transmitted over the internet, with 50 zettabytes in storage.
In this digital epoch, ANI serves as a critical tool to harness this vast wealth of information to enable humans to make better decisions in the industries of Healthcare, Energy and Finance. However, connecting ANI to the internet for real-time data consumption demands proper safeguards and rules to prevent detrimental outcomes.
Generative AI, like ChatGPT, relies on careful curation of substantial data volumes.
This technology, coupled with Subject Matter Experts’ (SME) insights, can streamline content creation, maintaining the SME’s unique voice. For example, an expert on Active Directory Federation Services (ADFS) could feed their previously created documents into an ANI. The ANI would then generate a new document, incorporating any recent updates from the internet, thereby enhancing productivity, and allowing the SME to focus on higher-level thinking.
The terms AGI and ASI often spring to mind when discussing “Artificial Intelligence”. AGI, or Artificial General Intelligence, represents an AI capable of performing any task a human can do, including passing the Turing test and exhibiting unique thought processes. However, achieving AGI remains a distant prospect, with new AI research models still under development.
ASI, or Artificial Super Intelligence, implies an AI level surpassing human intelligence in all domains. Although ASI is currently theoretical, futurist Ray Kurzweil predicts its emergence as early as 2045 in his book “The Singularity is Near“.
In essence, the future of AI is vast, stretching beyond popular perceptions.
As we continue to refine and expand these technologies, their potential to transform our world becomes increasingly evident. Generative ANI like ChatGTP4 can even re-phrase an article like this in a matter of seconds.