The other day, I used OpenAI’s new Whisper algorithm for the first time…

…Only to realize something very, very strange.

If you’ve not heard about Whisper, it’s OpenAI’s automatic speech recognition (ASR) system, and it’s significantly more accurate compared to something like Siri, which I usually use, or other kinds of technologies.

Anyway, for an upcoming Medium piece, I chose to narrate everything into the voice memo app on my iPhone, in preparation to have it transcribed by Whisper, while I was on a drive from my home to my cello lesson about 15 minutes away.

Whisper did it *almost* perfectly!

Whisper did it (almost!) perfectly, by the way.

Then, I uploaded this into ChatGPT just to format it but didn’t change any of the text.

This was perfect! It was well transcribed, everything looked good, and all that remained was for me to just post the thing, right?

…No.

Because you see, at this point, I started to wonder about a strange question that was starting to brew in my mind:

Was this text AI generated or was it human generated?

I asked several people this question, and almost all of them, said the same thing – it was AI assisted, but the text itself was primarily generated by a human.

To me, that makes a lot of sense because I recited it from my voice, and it wouldn’t make sense if it were considered to be AI generated unless what’s inside my head is not in fact a human brain, but rather some sort of super computer.

So I decided to just check with GPTZero just to be sure.

Here’s what GPTZero had to say, as it casually marked the parts of the my essay that it thought were AI-generated in a bright yellow.

I was a little shook.

Essentially, the majority of the text was determined to have been generated by AI.

At first, I thought about a couple of different possibilities. I wonder to myself – was it because I had put the text through ChatGPT? Could it have been that the text had been watermarked or modified in some way that allowed GPTZero to determine that it had been generated by AI?

That didn’t seem to make sense, particularly since OpenAI has not yet implemented watermarks in text . Still, the text definitely wasn’t modified in any way apart from the paragraphing or anything else of that nature.

Therefore, what was the only logical possibility here?

The only possibility is that according to GPTZero, I sounded like an AI.

This made me think quite a fair bit and it just so happened that in the gym I ran into my friend Sandy Clarke, with whom I ended up discussing the matter, (Sandy is wonderful and incredibly humble relative to what he’s achieved – check him out here!) and who suggested that perhaps artificial intelligence speech is just speech of a formal nature and to consider the speeches of JFK and Obama, so I decided to go right ahead and input JFK’s inaugural speech into GPTZero:

…Then I went on to input Obama’s inaugural speech:

… So did this mean that John F Kennedy and Obama were both advanced forms of artificial intelligence sent to planet earth to rule over the most advanced societies in the world, over which no normal human could have presided?

It would be funny if that were true.

At this point, I started to realize that the way that I spoke was just similar to the way that these people speak, which was similar to what GPTZero was identifying as AI-generated.

That’s not all that surprising, since my job is to help students learn how to write effectively, to assist them with their grammar, and also with the way that they use the English language.

But it made me start to wonder – when we interact with artificial intelligence, it’s a new type of interaction where we’re essentially just conversing with and reading from a tool that we are constantly interacting with all the time. Is it all that surprising that that could lead to language change on our parts, and therefore a shift in the way that we think and communicate?

It’s not necessarily going to be the case that humans end up fusing with machine parts, as some movies seem to suggest that we will, but certainly there are going to be changes in our culture as a result of the way that we interact with technology that perhaps aren’t immediately apparent at the outset; what are those changes going to be? It’s not immediately clear what the answer to those questions are.

It was definitely funny to think about this, though, because it leads them into all sorts of interesting questions about sentience, and also about the people that we communicate with – what if the people around us end up adopting artificial intelligence language patterns to the point that we are unable to distinguish the language that is used by artificial intelligence from the language that is used by human beings?

That might be one of the ways in which we become more machinelike as a species, or perhaps not — either way, it was pretty interesting to watch this happen and to ask myself about the ways in which I am being influenced by AI, because we often think of humans and AI as being distinct and different from one another, and that there are clear boundary lines that separate us…

But how are those boundary lines changing over time? The answer to that question is unclear to me.

Yep.

As we interact with AI, I suppose that we will start to talk a little bit more like AI.

As we move forward in this world, I expect that AI detectors will not really be a meaningful way in which we detect human beings — that our natural instincts of judgment and distinction may become just a little bit finer as we go through life.

On my part, I find it kind of funny that maybe the people reading this piece might think rightfully that I am an AI — a sentient AI, maybe — but an AI for all intents and purposes.

To that I say… Who knows if that could happen to you too?

Leave A Comment

Recommended Posts

Doc.new

Just discovered the doc.new shortcut, and it’s lifechanging.  All you do? Go to Chrome, and type in “doc.new” into the address bar, and poof – here you are, with a brand new Google Document. Why do I even know this? Because I use Google documents every day, and I like to make things just a little easier for myself so I don’t get the excuse of saying that I didn’t do things because they were too cumbersome or too difficult.  Here, I was trying to get a shortcut to create a new document and I was looking for the easiest possible way to do it – a way of enabling me to do things more easily, in more refined a fashion, in more simple a way to make things happen and develop. Docs.new is one of the most elegant things I’ve discovered this entire year, and it’s a shock that that realization came in nothing more than a single search for the shortcut and a single phrase typed into a keyboard. It makes me wonder how many other instances of this exist out there in our strange universe.

Some Thoughts on YouTube

Lately, I’ve become a lot more consistent with making YouTube content, but it’s not because of any sort of planning or anything – it’s because I’ve become a lot more stubborn, dogged, and just don’t really care as much what people think. Maybe it’s because I’ve gotten a little older now, maybe it’s because I no longer care, or maybe it was a skill issue – I won’t really know until I do my self-analysis, which I hope to do progressively as I compare my scripts to what I’ve done along the way, which I would like to do and hopefully will succeed at some time soon. Anyway, I thought this would be a fun post to think about what I’m putting out there and why, which kind of extends to the question of what I’m doing with social media anyway. But first… Why Even YouTube? YouTube to me is one of the best art forms that I have access to, and it’s one of the most enjoyable pastimes to me. It’s not even a pastime that I’m particularly good at, but it’s something that gives me meaning in a whole bunch of different ways because it’s enjoyable – something that blends together my feelings at any moment with that wish somehow to craft things for this world. You see, YouTube is about videos, and videos are an immersive experience and a recorded section of reality. The thing is (and we could go deep philosophical into this but this really isn’t the point of this blog post) videos don’t even have to be about the tangible and the everyday – they can just be selections or samplings of experiences that narrow down that experience into a single channel; a collection of moments seen, created, formed – a targeted crafting of reality that is very different from say, writing a blog post […]

Today’s Morning Reading

My morning began with the voice of David Brooks tearing down the elite class – it was a voice that I hadn’t heard for the longest of times, after procrastinating on replying a text message from someone for the longest of times. I think that it is worth a watch. Beyond what’s implied by the title of the video, Brooks discusses the evolution of merit, how the ‘elite’ was once defined and redefined in America as the Mayflower class transitioned into classes at Harvard and Stanford – how behaviors designated as desirable for our future leaders evolved through time as the generation took a turn and the world evolved. Midway through the video, I realized that I was distracted and thinking of something else – I began thinking about things that weren’t really related to what I was reading – but somehow through the pathway of internal reflections, a part of my conscience led me to read his “How The Ivy League Broke America” in its full 10871 word glory. Many thoughts went through my head at that point and still are at the moment – but Brooks expressed it better in that piece than I can, and I suggest that you read it. For what it’s worth though, here’s what I’ll say: His thoughts made me tap into an intuition that I’ve been having for a while – that intelligence isn’t really the primary determinant of life outcomes, and that there are other qualities and characteristics that I need as a person to continue pushing forward to have a fuller life, fuller existence, and everything else. Even now, my thoughts are evolving, and who I am as a person is changing – and it is fascinating to see that process take place, even if I’m not constantly watching every single detail of it – but that’s a story for another […]

Kamala Harris and the Overton Window 

When I saw the headline, my eyes widened.  “How Kamala Harris burned through 1.5 billion dollars in 15 weeks.”  I stared at my phone once, and I stared at it twice, as my disbelief grew.  Was this not NYT? Did they not just endorse Kamala with the force of an angry democratic tiger no less than two months ago? Was this real? As I thought through the implications, I saw my disbelief echoed in the comments that came along with it, the shock that filled my mind – the discomfiting revelation. Our world had transformed.  The New York Times is a paper that is unique amongst many others. First among equals in the world of newspaper journalism, its eminence has proven itself through the years and across eras as it shapes the way that the world thinks on a range of different issues, alongside its counterparts such as The Wall Street Journal, the Associated Press, and The Washington Post in the U.S., and on an international front, the BBC and Reuters in the United Kingdom, and Al Jazeera in the Middle East. There is an interesting adage that goes as follows: “When America sneezes, the rest of the world catches a cold”. To that I say, whatever the New York Times publishes, the world garnishes as the realm of acceptable discourse unfolds, an entire communicating planet paying homage to the one of the dominant media voices in the United States of America. …Which leads to my question.  What does it mean when NYT tells us about “How Kamala Harris Burnt Through $15 billion in 15 weeks”? The article I would like to write is not an article about campaign spending, and neither is it a piece to point out flaws or discrepancies in Kamala’s campaign: The first would be far too boring, and the second would land me into polemics in […]

Becoming Less Verbose

Children learn all sorts of things through lots of different ways, whether it’s school, whether it’s at home, whether from the local tutor or in the church. But you know, one of my favorite learning methods is the fight. Now I know it sounds bad and I certainly don’t mean UFC grappling and punching hijinks, but I will admit that I can be combative at times, and I’m not particularly afraid of fighting with words, which often kind of makes things worse, but then c’est la vie, it is my personality and the way that I personally get to truth one argument and one debate at a time, hopefully listening a little more than I speak but then making sure to clarify whatever I feel that we cannot say we know. I recently had a fight with someone in Mensa International, and I chose to block that person. I mean, fair dice – Mensa International, which is the main Facebook group of Mensa as an organization determined by the Mensa International Board of Directors (IBD) is the single largest gathering of Mensans, who – while I love many of them – can also be some of the most irritating people in the world. Anyway, we had a dispute about of all things, Kamala Harris, in relation to this exciting headline: …Which has attracted some rather interesting comments: This was a pretty interesting topic (which I will write about) because of how it showcases a shift in the Overton Window while at the same time showcasing media bias in a range of different ways. Anyway, this individual had commented. I’d had a conflict with her before on account of her attempting to use her background to win an argument once (a clear pet peeve of mine which I might talk about a little more later on) but thought that okay, I can’t […]

The Body is the Hardware, The Mind is the Software

The analogy was interesting when I heard it first, and it remains interesting now because it resonates with me on at least a couple of different levels. Our bodies, the physical parts of us, are basically analogous to the hardware of a computer, running along with different parts here and there – upgradable, we can improve them by increasing the quality of the resources that go into them; improvable through good maintenance, we can exercise, sleep well, and do all sorts of other things to improve the hygiene on that front. Our minds, on the other hand, are the software – the programming that decides how we interact, think, solve problems in specific situations; the algorithms and little decisions that decide how we react to different scenarios and confronting different situations, whether it comes to talking to girls, investing, selling, marketing, or doing business with others. It is nice to think that the mind is upgradeable, and that somehow you can improve yourself through an act of willpower by learning certain things. Through sitting down and unlocking the secrets of the universe one after another, through a mixture of magic and also destiny. But who’s to say exactly how that should happen? Sorry, that’s a silly question. The answer is that it’s you.