 |
| Hype Masks Real Harms Threatening Our Present, Not an Imaginary Future |
Forget apocalyptic robots. The genuine dangers of artificial intelligence lurk not in science fiction but in our daily lives, from biased algorithms to stolen data and fake news factories.
While headlines scream about AI wiping out humanity, a far more insidious reality unfolds: existing AI tools already harm individuals and society across various domains. This fearmongering about fantastical futures deflects attention from the present perils we face right now.
Discriminatory housing algorithms, biased criminal justice systems, and the spread of misinformation in diverse languages are just a few examples. Algorithmic management tools quietly steal wages from workers, and the trend is growing.
Corporate AI labs benefit from this distraction. They churn out "research" filled with buzzwords like "existential risk" to divert attention from real issues and gain regulatory leeway for their unproven technologies.
Instead of succumbing to AI hype, we must listen to scholars and activists analyzing its immediate impacts. Their work unveils the dangers embedded in today's technology.
The term "AI" is ambiguous, encompassing various fields and techniques. In practical terms, it often refers to pattern-matching algorithms trained on massive data sets. But in marketing and startup pitches, it becomes a magic wand promising miracles.
Large language models like ChatGPT generate remarkably fluent text, yet lack true understanding. They're like elaborate Magic 8 Balls, mirroring the biases encoded in their training data – bigotry, misinformation, and all. This undermines trust in information, making it harder to discern truth.
Proponents tout these machines as answers to various societal needs, from education to healthcare. But they exploit workers by stealing training data and subjecting others to repetitive, low-paid labor to "clean up" biased outputs.
Employers use automation to cut costs, displacing workers and then rehiring them at lower wages to fix the AI's mistakes. This fuels exploitation and undercuts living standards, as seen in recent Hollywood strikes.
AI policy needs a foundation of accurate research, yet much is funded by corporations or biased labs. This "junk science" hides behind secrecy, exaggerates capabilities, and uses flawed evaluation methods. Examples include Microsoft's claim of "intelligence" in its text-synthesizer or OpenAI's unsubstantiated boasts about problem-solving abilities.
"AI doomsdayers" use this junk science to focus on fantastical threats, distracting from the real harms at hand.
Policymakers must prioritize solid research on AI's immediate risks, including societal impacts like disempowering the poor and intensifying biased policing. Only then can we develop sound policies to minimize harm and ensure responsible development of AI technology.
Let's move beyond the AI hype and address the real issues that threaten our present, not some distant imaginary future.
