I’ve been doing quite a bit of reading recently to prepare for BiZ Gear Up on the 24th, and part of that led me to read a bit more about the early history of chatbots – I hope you’ll enjoy this one!
We take a brief moment to move away from the hype that is OpenAI‘s ChatGPT, and take a brief intermission as we make a small trip back in time.
Picture this: it’s the year 1966, and a computer scientist at the Massachusetts Institute of Technology‘s Artificial Intelligence Laboratory named Joseph Weizenbaum has just created something remarkable – the world’s very first chatbot: ELIZA.
Image credit: Public domain via Wikimedia Commons
Just consider this example of a conversation from Norbert Landsteiner’s 2005 implementation of ELIZA, and you can see what it was capable of.
ELIZA was designed to simulate conversation by responding to typed text with pre-programmed phrases and questions.
But what made ELIZA so special was that it was programmed to mimic the conversational style of a therapist, in particular a Rogerian therapist.
Users could “talk” to ELIZA about their problems and concerns, and the chatbot would respond with empathetic and non-judgmental phrases like “Tell me more about that” or “How does that make you feel?”
It wasn’t just a simple question-and-answer program – it was designed to provide a sense of emotional support and understanding that reflects interestingly on the ways that people derive comfort from self-affirmation.
Weizenbaum didn’t intend for the chatbot to be taken very seriously, calling it a “parody” in his 1976 book “Computer Power and Human Reason”… But the way that the chatbot was received was far from just a parody.
The response to ELIZA was overwhelming.
People were fascinated by this new technology that could seemingly understand and respond to their thoughts and emotions, and the program quickly gained popularity as people tried the chatbot.
But perhaps what’s most remarkable about ELIZA is that it wasn’t just a novelty. Weizenbaum’s creation laid the foundation for decades of research in the field of natural language processing and artificial intelligence.
ELIZA was the first chatbot, but it certainly wouldn’t be the last – and its legacy lives on in the many conversational AI programs we use today, in our Bings, Bards, ChatGPTs, Claudes, and the many more that exist and will exist today.
Today, it exploded and hit 1k, and it shows no sign of stopping; the hype is real!!
Why?
A mixture of Manglish, an anonymous dude who prompted OpenAI’s ChatGPT with the legendary Manglish prompt, the combined efforts of Kenneth Yu Kern San and Jornes Sim, and, I guess, a huge, huge dose of the wonders of ChatGPT x)
My own small contribution (lol!)
I guess I’ve evolved into the comments guy – take a look here if you want đ
Ah, what can I say?
It’s truly fascinating to be at the beginning stage of a revolution.
First off, congrats and creds to Pang for being infinitely better than I am at managing communities at the moment – something I’m definitely looking forward to learning more in the days ahead!!
Second off, I care a lot about the deeper significance of things – and I’m incredibly glad that this is one of the many things that’s starting off AI on the right foot in Malaysia, my home country – where this will go and what will happen I have no idea, but really look forward to watching what the world’s going to bring đ
Amongst other things, I guess it’s brought a ChatGPT Plus subscription for which now, in order to sign up, you’ve got to join a waitlist.
Can’t wait to see what’s next, and thank you based Sam Altman and OpenAI for creating this gift to humanity đ
âUpload your essay into Turnitin by 11:59pm on Thursday night? You meant start the essay at 11:57 then submit it at 11:58, am I right?â â Gigachad ChatGPT student.
We begin our discussion by discussing the sweet smell of plagiarism.
It wafts in the air as educators run around like headless chickens, looking here, looking there as they flip through oddly good essays with panicked expressions.
âWas this AI-generated, Bobby?!â says a hapless teacher, staring at a piece of paper that seems curiously bereft of grammatical errors, suspecting that Bobby could never have created something of this caliber.
âNo teacher, I just became smart!â Bobby cries, running off into the sunset because he is sad, he is going to become a member of an emo boyband, and he doesn’t want to admit that he generated his homework with ChatGPT.
generated by Midjourney, if that wasnât obvious
This smell casts fear and trepidation over every single part of our education system, for it threatens to break it; after all, education is special and it is meant to be sacrosanct â after all, is it not the very same system that is designed to teach humans facts and knowledge and above all, to communicate and collaborate to solve the problems of our era with intelligence, initiative, and drive?
Itâs unsurprising that the world of education has flipped out over ChatGPT, because artificial intelligence opens up the very real possibility that schools may be unable to detect it.
Fun and games, right? Itâs just a bunch of kids cheating on assignments with artificial intelligence? Itâs not going to affect the older generation?
As it turns out, no â thatâs not the case. Iâll explain why later.
But before that, letâs talk a bit about the part of our education system that AI is threatening: Essay-writing.
If students simply choose to let their work be completed by artificial intelligence and forget all else, that just means that theyâve forgone the education that theyâre supposed to have received, thereby crippling them by an act of personal choice, rightâŠ?
But each of us has been a student, and if we have children, our children either will be or have been students too; there is a deep emotional connection that stretches across the entire world when it comes to this.
Therefore, when Princeton University CS student Edward Tian swooped in to offer a solution,itâs not all that surprising that the world flipped.
Enter GPTZero.
Humans deserve the truth. A noble statement and a very bold one for a plagiarism detector, but something thatâs a little deeper than most of us would probably imagine.
But consider this.
Not everyone who uses AI text is cheating in the sense of doing something that they are not supposed to and thereby violating rules, therefore the word âplagiarism detectorâ doesnât quite or always apply here.
This algorithm, as with other algorithms that attempt to detect AI-generated text, is not just a plagiarism detector that merely serves to catch students in petty acts of cheating âit is an AI detector.
An AI Detector At Work.
So how does it work?
GPTZero assigns a likelihood that a particular text is generated by AI by using two measures:
Perplexity, and Burstiness.
Essentially, in more human language than that which was presented on GPTZeroâs website, GPTZero says thatâŠ
The less random the text (its âperplexityâ), the more likely it was generated by an AI.
The less that randomness changes throughout time (its âburstinessâ), the more likely that the text was generated by AI.
Anyway, GPTZero gives each text a score for perplexity and burstiness, and from there, outputs a probability that given sentences of a text were generated by AI, highlighting the relevant sentences, and easily displays the result to the user.
Alright, sounds great!
Does GPTZero deserve the hype, though?
âŠDoes this actually work?
Letâs try it with this pleasant and AI-generated text that is exactly about the importance of hype (lol).
Thatâs 100% AI-generated and we know that as fact.
âŠWould we know if we didnât see it in the ChatGPT terminal window, though?
âŠOkay, letâs not think about that.
Down the hatchâŠ
âŠAnd boom.
As we can see, GPTZero, humanityâs champion, managed to identify that the text that we had generated was written by AI.
So nope, GPTZero canât detect rewritten texts that were generated with AI â which it should be able to if it truly is an *AI* detector in the best sense â and which in turn suggests that the way that itâs been operationalized has yet to allow it to be the bastion protecting humanity from the incursion of robots into our lives.
Itâs not that GPTZero â or even OpenAIâs own AI Text Reviewer, amongst a whole panoply of different AI detectors â are bad or poorly operationalized, by any means. Rather, itâs that the operationalization is supremely difficult because the task is punishingly hard, and that we are unlikely to have a tool that can detect AI-generated text 100% unless we perform watermarking (MIT Technology Review) and we would have to use multiple algorithms to be able to detect text, or come up with alternate measures to do so.
An Arms Race between AI Large Language Models (LLMS) and AI Detectors â and why you should care (even if youâre not a student).
As Iâve mentioned, there is an arms race at hand between AI Large Language Models (LLMs) like ChatGPT, and AI detectors like GPTZero, the consequence of which is likely that the two will compete with one another and each will make progress in its own way, progressing the direction of both technologies forward.
Personally, I think that AI detectors are fighting a losing battle against LLMs for many reasons, but let me not put the cart before the horse â it is a battle to watch, not to predict the outcome of before itâs even begun.
Implications of this strange war:
But why should you care about any of this if youâre not a student? Itâs not like youâre going to be looking at essays constantly, right?
Letâs take aside the fact that youâre reading a blog post right now, and letâs also move away purely from the plagiarized essay bit that weâve been thinking about, as we gravitate towards thinking about how ChatGPT is alanguage model.
Itâs a good bet that you use language everywhere in your life, business, and relationships with other people in order to communicate, coordinate, and everything else.
When we go around on the Internet, itâs not always immediately evident what was AI generated, what was generated by a human or, for that matter, what was inspired by an AI and later followed through by a human.
The whole reason we need something like a plagiarism detector is that we may not even be sure that a particular piece of language (which we most often experience in the form of text) is AI-generated with our own eyes and minds, to the point that we need to literally rely upon statistical patterns in order to evaluate some thing that we are looking at directly in front of us, thereby recruiting our brains as we evaluate the entirety of an output.
The problem isâŠ
Language doesnât just exist as text.
Language exists as text, yes, but also as speech. Moreover, speech and text are easily convertible to one another â and we know very well what ChatGPT is doing: Generating text.
We now know that there are Text To Speech (TTS) models that generate speech from text. Theyâre not necessarily all great, but thatâs besides the point â it presages the translation from text into voice.
Think about it.
If the voices that are generated by AI become sufficiently realistic-sounding and their intonations (VALL-E, is that you?), how might you know that these voices arenât real unless there are severe model safeguards that impede the models from functioning as they are supposed to?
Now combine that indistinguishable voice with sophisticated ChatGPT output that can evade any AI detector and in turn may, depending on the features that end up developing, evade your own capacity to tell whether you are even interacting with a human or not.
How would that play out in the metaverse?
How would that play out in the real world, over the phone?
How would you ever know whether anyone that youâre interacting with is real or not? Whether they are sentient?
The battle between AI and AI Detectors is not just a battle over the difference between an A grade and a C grade.
Itâs a battle over a future where whatâs at stake is identifying what even qualifies as human.
Do you ever feel like you might have gotten yourself in something a little bigger than you’d imagined was possible?
Excited to announce that I’ll be speaking about AI for the “How AI Tech Will Disrupt Businesses” panel on the 24th of February! Thank you Vulcan Post for the feature and MrMoney TV x Entrepreneurs and Startups Malaysia for the invitation!
You’ll be able to meet me there directly and hear me talk about the ways AI is going to change businesses around the world alongside my fellow panelist, Richard Ker!
If you’ve not heard of Richard, the man is a legend at creating incredible infographics and marketing, and I respect both his trite observations and the value that he’s created for literally thousands of people throughout Malaysia and far beyond; the man is a true blue digital authority, If you’re looking for something specific, feel free to check out this article that he’s written on Facebook, amongst other things; the man is everywhere!
In other words, what does this mean?
It means I need to level up!
This conference is something that I’m truly honored to be a part of, and a wonderful opportunity to learn from many incredible minds that I won’t be missing by any stretch of the imagination.
As we speak, I’m preparing for with all my might at the moment even as I read and learn more about artificial intelligence, building up that reading habit again thoughtfully documented by my dear friend Sandy Clarke and that I’ll make sure to work towards in the days ahead as I build this platform.
Meanwhile, if you haven’t already, please feel free to join Artificial Intelligence Malaysia! I’ve had some pretty wild conversations in the past day or so, and it would be great to add a diversity of voices to the group especially if you’re really interested in AI and everything that it has to offer đ
In preparation for that, know that I’ve been reading extensively and creating lots of other content as well because I know that anything I have to make this worth your time, and will do my very best to do so.
