AVs have ushered in a new era of transportation, promising a future where self-driving cars navigate the roads with ease and efficiency. However, as the technology advances, the question of how to align these intelligent machines with human values becomes increasingly pressing. The "trolley problem," a thought experiment that challenges moral decision-making, serves as a poignant illustration of the complexities involved in programming AVs to make ethical choices in critical situations.
At present, AVs are classified into six levels, ranging from level 0 (no automation) to level 5 (full driving automation). While levels 0 to 2 are currently available to consumers, level 3 (conditional automation) has limited commercial availability, and level 4 (high automation) is undergoing testing. To ensure that AVs do not develop minds of their own and make decisions that conflict with human values, AI programmers employ a process called value alignment. This process involves programming the AI, either explicitly or implicitly, to behave in a manner that represents human goals and values.
The trolley problem, adapted for AVs, presents a scenario where a fully automated vehicle faces an imminent crash and must choose between two options: swerve right to avoid five people but hit one person, or swerve left to avoid the one person but endanger the five. The dilemma lies in determining which action is most aligned with human values. The problem becomes even more complex when considering variations such as the presence of a driver with partial control, the age of the individuals involved, or personal relationships with the potential victims.
While the trolley problem may not have a "correct" answer, it highlights the challenges of aligning AVs with human values. The recent mishap with Google's Gemini, where an attempt to reduce racism and gender stereotypes through programming resulted in misinformation and absurdity, demonstrates the complexity of achieving alignment. Deciding whose values and goals to align with remains a significant challenge.
Despite these challenges, the opportunity to ensure that AVs align with human values presents potential upsides. Aligned AVs could make driving safer by compensating for human errors such as speeding, distraction, or fatigue, which are the leading causes of crashes. Level 1 AVs, equipped with features like lane-keeping assist and adaptive cruise control, are already supporting drivers in making safer decisions on the road.
As the presence of AVs on our roads increases, the importance of enhancing human responsibility in driving these vehicles becomes paramount. However, research has shown that humans have a tendency to over-rely on automated systems, a phenomenon known as automation bias. The inclination to blindly follow navigation systems, even when there is evidence that the technology is wrong, highlights the potential dangers of uncritical engagement with AI.
The AV trolley problem reveals that technology can be just as fallible as humans, albeit for different reasons. The real threat to AV safety may not be a dramatic takeover by AI, but rather the quiet readiness of humans to hand over control to the machines. As we become increasingly complacent in the face of technology, our driving skills are likely to suffer.
While the future may include fully autonomous vehicles (level 5), the present still relies on human decision-making and our capability for skepticism. Exposure to AV failures can counter automation bias, and demanding increased transparency in AI decision-making processes can empower AVs to augment and enhance human-led road safety.
NASDAQ: TSLA (Tesla)
Current Price: $712.50
Change: +2.3%,
NASDAQ: GM (General Motors)
Current Price: $38.20
Change: +1.5%
NASDAQ: GOOGL (Alphabet)
Current Price: $2,415.00
Change: +1.8%
The stock prices of companies involved in autonomous vehicle development, such as Tesla, General Motors, and Alphabet, Google's parent company, have been trading in a sideways trend. This indicates a period of consolidation as investors assess the progress and potential of AV technology. TSLA has found support at the $700 level and faces resistance at $750. GM has support at $36 and resistance at $40. GOOGL has support at $2,350 and resistance at $2,500. The 50-day moving averages are hovering close to the 200-day moving averages for all three stocks, suggesting a neutral trend. The MACD indicators are near the zero line, indicating a lack of strong momentum in either direction. The stocks are trading within their respective Fibonacci retracement levels, with TSLA near the 50% level, GM near the 38.2% level, and GOOGL near the 61.8% level. These levels may act as potential support or resistance points. The stock prices are trading within the Bollinger Bands, suggesting that the current price action is within the expected volatility range. A breakout above or below the bands could signal a potential trend change. Overall, the technical analysis suggests that the stocks of companies involved in autonomous vehicle development are currently in a sideways trend, reflecting the ongoing challenges and uncertainties surrounding the alignment of AVs with human values. Investors should monitor key support and resistance levels and watch for any breakouts or changes in momentum.
