Regardless of the type of emergency, communication is a vital thread between life and death. However, reaching diverse groups isn’t easy, and if we as a society learned anything from past natural disasters, it’s that misinformation, delays, and other challenges often creep up at the most inopportune moments.
This isn’t a business environment where delays end in losing company money. The stakes are much higher, so any issues, sadly, result in loss of life, spread of fear and panic, and oftentimes diminished public trust.
Yet, we may be close to remediating this issue, or at least, lessening its impact by using AI. Instant context-aware alerts have the potential to reach targeted populations instantly, thus bridging any communication gaps that can end in disastrous consequences.
High-stakes game of telephone
Before I analyzed the situation closely, I assumed that the challenges in emergency communications were purely the result of panic. While this is partially true, since human nature is a massive contributor to the preventable loss of life, the 2023 Maui fires made it clear that the traditional systems and the ingrained culture were actually the key factors in the catastrophe.
The problem is that traditional alert systems are dependent on existing infrastructures. A single disruption of internet connections can significantly delay critical warnings. Sadly, it doesn’t end there. Delays also happen due to our reliance on press releases, manual input, and something as trivial as bloated bureaucracy. Of course, this is unacceptable in rapid crises where even a slight hold-up can lead to death.
The telephone game gets even more fatal when language barriers occur. Since large areas are now diverse, many people who don’t speak the dominant language are often sadly left in the dark. Even if they do receive instructions, haphazard translation can also contribute to confusion, which can be deadly in such a scenario.
Social media, for all its convenience, has also had an irreversible impact on our emergency protocols. Words on these platforms move fast and often leave official updates in the dust. In moments of chaos where every second matters, it’s defeating to see people falling prey to misleading, false, or even malicious posts.
As a result of social media, people may end up losing trust in official instructions, as we’ve seen many times, like during the COVID-19 pandemic.
How AI can cut through the noise
Because we’re dealing with such a complex communication issue where multiple links could crack in an instant, AI is the only way to strengthen the entire chain, so to speak.
For starters, it can almost instantaneously process data from various sources and deliver alerts faster than the existing methods. Speed saves lives, so AI may be able to allow emergency response teams to reach people long before the situation escalates.
Also, AI alerts can erase (or at least bridge) language barriers. The technology can efficiently and accurately localize messages in real time, strengthening clarity and minimizing the spread of panic.
Challenges and ethics of AI in emergency communications
Policymakers aren’t as optimistic as I am about newer technologies. They may criticize AI systems for their propensity for errors, which is a sentiment even some proponents parrot: “AI should never be used in life-or-death situations.”
This take lacks nuance. Why?
Simple – AI systems are only as accurate as their design. It’s possible to train crisis communication systems only on verified datasets, conduct thorough testing, and most importantly, monitor the results to preserve reliability.
We humans also tend to exaggerate our mental prowess, so another common argument against using AI in emergency communication is that human decision-making is simply more reliable.
Any way you slice it, overreliance on AI is generally a subpar approach, whatever the context. Human operators should still retain full control, with AI only serving to augment their abilities by providing fast analysis and distributing the alerts. Such a system must have multiple levels of ethical controls and verification layers in place to minimize the risk of spreading misleading or false information.
At face value, developing AI for use in emergencies is not only difficult but also costly to implement. Nonetheless, since the end goal is saving lives and minimizing chaos in catastrophes, developing and implementing such a system is more than worth the cost.
Preserving life and trust
Relevant government agencies have an opportunity to preserve life and property by using AI in emergency situations. Still, reaching the pinnacle of this goal will require close collaboration between all the key players – from government agencies themselves to emergency responders and AI developers.
An ethical and responsible approach to such a crucial piece of technology will, by extension, strengthen public trust.
The only question is if we’re ready to overlook costs and potential implementation challenges to protect something most sacred – human life.
