FAIK: A Practical Guide to Living in a World of Deepfakes, Disinformation, and AI-Generated Deceptions by Perry Carpenter

Author:Perry Carpenter
Language: eng
Format: azw3
Tags: Network Security, Computers, General, internet, Networking, Security, Online Safety & Privacy
ISBN: 9781394299898
Publisher: Wiley
Published: 2024-08-20T07:00:00+00:00

Into the Exploitation Zone

As I said earlier, deception has been part of the human condition since the beginning. What changes is the ease and scale at which we can fool each other and the global impact a single deception can have. Recent advancements across multiple areas have pushed us well past the crossover point where technological innovation is exponentially outpacing human adaptability, leaving most of society in the exploitation zone.*

We’ve been feeling this tension for years.

With the pace of innovation accelerating daily, that gap between human adaptability and technological progress keeps getting wider. And the current proliferation of generative AI-related research and advances has the effect of throwing gas on an already blazing fire. So, before we finish this section of the book, let’s hop into our version of the Scooby Doo Mystery Machine. We’ll take a brief tour of the exploitation zone to get an idea of where we are now and what’s coming. Along the way, I’ll unmask a few AI-generated threats and dangers we can expect to see as we navigate this expanding divide.

To set a bit of context, let me introduce you to what’s being called the jagged technological frontier.1 This jagged frontier concept has important implications for how generative AI can help virtually anyone level up their abilities. Some activities that seem complex for humans are quite easy for current AI models, while other tasks that appear simple remain challenging for AI. This creates a jagged frontier where AI suddenly becomes highly capable at certain tasks while still struggling with others.

To see how AI impacts knowledge work, researchers ran an experiment with hundreds of consultants.2 They split the consultants into two groups: one had access to GPT-4 (the same model available for free on Bing or through OpenAI’s $20-per-month plan), and the other didn’t. No fancy tweaks. No special settings—just standard GPT-4.†

Researchers gave the consultants tasks from a fictional shoe company, covering creative ideas, analysis, marketing, and persuasive writing. They even checked with a real shoe company exec to make sure these tasks were legit.

The results? The consultants who partnered with AI crushed it. They finished tasks faster, completed more tasks overall, and produced higher-quality work. This was true whether they got a quick intro to AI or not. And both human and AI graders agreed on the quality of work produced.

AI turned out to act as an equalizer across skill levels (see Figure 7.1). Consultants at the bottom half of their skillset saw huge improvements—a 43 percent jump when using AI. And that was enough for them to outperform top-half participants who didn’t have access to AI. The top-half participants saw an impressive but not as dramatic improvement of 17 percent. Further, when participants across both groups had access to AI, their scores were extremely close. The experts still outperformed those in the lower ranks, but only slightly.

It’s like how steam shovels changed mining—making everyone’s digging skills almost irrelevant.

Here’s the takeaway I want you to consider: generative AI brings with it the ability to level the playing field between those who are unskilled and those who are highly skilled.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.