AI has become something of a bogeyman and it is not helpful to view the technology in this way. It is indeed often hyped up junk, as when Apple’s AI helpfully prioritizes spam email as ‘urgent’ for me. Yet as with any other technology it brings change and benefits and problems. I use AI-enabled tools every day. It concerns me that a negative narrative has emerged, which purely focuses on AI as a threat to individual human creativity and to the environment. This is a one-sided and rather hoary and dated critique that rests largely on an assumption that we cannot possibly integrate these technologies into our lives.Techno fear is as bad as techno hype and just because it feels good doesn’t mean you should do it.
My existence would be much diminished without AI tools. I am unable to use a keyboard or other input device. Most of my voice has gone. The only dictation support that works is an AI driven one called MacWhisper, which uses the Whisper LLM. This is the only transcription software that can make sense of my much diminished voice. It is far superior to the solutions offered by others. And to be clear, it is actually the commercial software that is best.You know that horrible integration that Apple does and that the EU wants to end and loads of tech critics complain about. The one where everything works together. Well, it’s the best thing I can have. Using macOS and an iPhone, I can take advantage of my existing contacts and software and integrate accessible workarounds and solutions into my existing workflow. I can use iPhone mirroring to access my phone directly from the desktop.This means I can continue to work and communicate with friends in a more seamless way.
This is far better than the clunky, awkward and disheartening solutions offered as part of specialist support for the disabled. On the one hand you need to have reliable solutions which necessarily are not going to be as pretty. On the other these all kind of assume that you’ve already given up on life. As an example, there is a system that will allow me to make calls using a head mouse tracker. But, you have to type in each phone number manually, like it was 1998 or something. I literally cannot remember the last time I manually entered a phone number. There is no possibility of integration with my existing, well, existence, my existing way of working, communicating with friends and loved ones. It’s just one example. The other solutions being offered are much the same. I can laboriously type out an email in a Windows pad. Or I can use my existing computer, which I can use a head tracking mouse with, and use voice control and dictation to send an email that way, in a way that I need to because I’m a professional still. This means I can continue to work and communicate in a seamless way. I can also control my phone directly from the desktop.Despite my situation, I am still writing books and papers. I am still making plans with friends and family.
As a researcher and a teacher, I have a complex workflow built up over many years that uses customised tools and complex data sets. I need these kind of data, this kind of information, this kind of multiple tools at my fingertips in a way I can use for my work and in fact for my pleasure. It is absolutely infuriating to be told that I can switch on my TV with the system as if that’s all I should have as a disabled person. My lot in life is just to sit in front of the telly all day. Well I can already put my Apple TV on with my Apple computer or my Apple iPad and using that I can watch all kinds of TV from around the world which is what I do. Everything about these tools is focused on the lowest common denominator passive consumer. That’s not who I am, or anyone is. We are still in our lives. Likewise, a new era of domestic robotics cannot come soon enough.
If we do not highlight these positive use cases for AI, then they will be regulated out of existence or their development will be limited and I will be a lot worse off and my life will be much reduced.
More generally, I find the way that AI is used as this kind of scare figure quite frustrating. It blocks analysis and evaluation. For example, the University proposes for its ethical investment strategy it will not invest in companies that produce AI-enabled weaponry, As if this is somehow more nefarious than just plain old artillery shells and such. No explanation is given as to why AI weaponry is so much worse than just lobbing a RPG at someone.. If an ally like Ukraine is using AI to target incoming missiles to support its air defence and for its own targeting, why is that a bad thing? It might actually save lives if it improves their air defence capability and makes their targeting more effective. It might shorten the war and in any case allies should have access to the means to defend themselves against one of the greatest threats to civilization of our time. If Ukraine is able to use autonomous weaponry to make up for its manpower deficiencies, then I do not see that as being a problem.
The distinction between AI and non-AI is dissolving anyway. For example, artillery shells can be networked to work cooperatively.
That does not mean that AI is always the right solution or even a working solution. It still needs to be treated with critical reasoning. There are plenty of sins to recount. It continues to occupy the place that data did about ten years ago in the minds of venture capitalists, as some magic dust to sprinkle upon a business plan. It will never understand irony, which makes parsing British communications very difficult. A lot of these sins, however, lie in the credulity of the beholder. For me, the bigger problem lies in the human willingness to give up our autonomy, either to the group or to a convenient technology or to the state. AI is just one insidious way in which this can happen, and people can farm out significant choices to a system that allows them to pretend they have no choice at all. But there are many other systems that do that, that we need to be alert to.
