A Note About AI
I went down a rabbit hole last summer reading about Chinese tech surveillance and the use of Artificial Intelligence (AI) and data mining to monitor and control their citizens.
I was a casualty of the AI mass tech layoffs and hiring freezes during my software engineer days.
I got heat for an essay in graduate school because the way I style-formatted a quote-block was flagged by an AI grading software as ‘plagiarism’ before my professor reviewed and overrode it.
I’ve been Facebook-jailed because the Facebook AI that screened my post for censorship didn’t understand human language nuance after I posted about a national battlefield.
Trust me. Between my nation’s enemy using AI, to losing my job, to questioning my academic integrity, to watching censorship go wrong… I’m well aware of how destructive this new technology is. I’ve personally felt it.
Just like how the internet and iPhone changed the world, the generation that grows up familiar with AI will behave and think wildly different than the ones who didn’t, the same way there exists a divide between folks who had iPads at age 5 and folks who lived pre-internet.
The ones who had the money and foresight figured out how to influence AI and own all the data marketed AI with glee, while the rest of us looked on with skepticism and fear.
And you’re not wrong for your fear and skepticism. But to only perceive AI with fear and skepticism is to ignore the good it can do for your local community, too. AI is a lot of job-taking and privacy-data-controversy. It is that. And it’s also more.
I will talk about two AI-focused companies that are doing what I call ‘good’ AI work.
DeepLook Medical
Firstly, DeepLook Medical, a company based out of New York that is focused on AI and women’s healthcare and radiology. For the last decade, they have developed an AI-software that will overlay on existing medical imaging tools. Not only will this software render higher-resolution graphics of a patient’s cancer mass or uterine polyps, but it will also render color and temperature variants and size measurements of that specific region. This full-colored, high data report can then use AI to find other anonymous patient reports who match in body morphology and mass sizes to calculate diagnosis, risk, and offer treatment plans for that patient. This report is evaluated by the patient’s radiologist to quickly discuss next steps.
This entire process cuts screening, diagnosis, and treatment time. So long multiple long-distance medical trips and multiple screenings and long wait times, the error rate on this software is less than 2%. So much of the pain of medical situations is in the long days of not knowing, not having a plan, not having the data or enough information, or having to travel long distances and spend money on travel when you wish you could put it all to treatment. That part of medical mental health issues is about to change. They are not yet approved for FDA, but I created the marketing website for them and am excited to see their technology develop and advance.
Safety
The second company is Omnilert, a New England company focused on harm reduction. Their AI software ran as an overlay on existing surveillance camera footage and was trained to identify weapons-ready-to-fire to prevent mass shootings. If a gun was holstered, it didn’t alert the school or hospital security team. However, if a firearm was in someone’s hand and ready to fire, the AI software would detect the gun make and model so the security team could execute safety protocols with all of this information readily accessible for the emergency responders. Our average time to scan, report, and prompt security teams to call emergency responders was less than 75 seconds from the first time the camera caught the armed individual. While this doesn’t prevent mass shootings, it does prevent lives from being lost. I served as a designer for the software, and I look forward to the day the regulatory processes pass so it can be used to save lives.
Artificial Intelligence, like the internet or like a firearm, is a tool. In the wrong hands, we’ve got ourselves a problem. A big problem. A China-sized problem. In the right hands, we’ve got ourselves solutions to medical diagnostics and active shooter harm mitigation.
This AI thing is powerful. It’s easy to focus solely on the bad, and ignore the benefits. When we cower from technology, when we decide to be afraid and ignore it, we will paint ourselves into a corner while the rest of the world flies right by us. Be diligent, be aware, but also be open-minded.
