hAIters’ dictionAIry
A small fun evening project that ended up much more fun than expected!
It all started with another semi-informed debate about AI-good-or-bad. So I decided to write the “hAIters’ dictionAIry” — proper scientific ontology for the observed phenomena during these debates. I started with the first 3 entries (kinda random, I know):
1. **rAIcist** — A person who tolerates certain flaws in humans, such as biases or inconsistencies, yet becomes critical when AI exhibits similar, albeit lesser, behaviours. This includes holding onto outdated knowledge, inconsistency, hallucinating, propagating hearsay, parroting (social media) instead of understanding, or engaging in unsafe practices, reckless driving, etc.
2. **rAInman** — Inspired by the film “Rain Man,” this term describes the hidden genius within AI, akin to Dustin Hoffman’s character, who, despite appearing limited, displays exceptional abilities with the right prompts. Recent studies have shown that with minimal guidance, AI can engage in sophisticated reasoning and planning, showcasing its latent talents when approached with care and understanding (Chain of Thought, ReAct, Self Discover, Voyager).
3. **planet of the AIpes** — A doomsday scenario in which the AIpes take control of the planet and us. Opening scene: Siri, “NLU” grammars, LSTM, Transformers, ChatGPT, GenAI, they learn to use tools, they throw a langchAIn into space — BOOM!.
I was suddenly struck with genius / laziness, and decided to ask ChatGPT for help. TBH, I had low expectations*.
*Another term came up as I was writing this: “low expectAItions — having low expectations from AI, despite the fact that it consistently blows our mind.“
Anyhow, back to the story: I gave the above 3 examples to ChatGPT 4 and asked it to suggest 10 more. To temper my low-expectAItions” ©, I intentionally used a very simple prompt with very few clues or guidance:
“Here is an idea I had with some examples. Note that there are no spelling errors — it’s all intended, and lowercase vs uppercase matters. Can you give 10 more examples: < 3 examples>”
This is what it came back with:
1. **AIgonize** — The process of worrying excessively about AI taking over jobs, society, and the world, despite AI being a tool designed to assist and enhance human capabilities. E.g., People who AIgonize often ignore the collaborative potential of AI in creative and problem-solving tasks.
2. **deceptAIon** — The act of blaming AI for being deceptive or manipulative, when in reality, it’s the input or the bias in the data provided by humans that leads to such outcomes. E.g., An AI that generates misleading information is often just mirroring the deceptAIon present in its training data.
3. **innovAIte** — The act of creating groundbreaking and innovative solutions through the use of AI, often overlooked by critics who fail to see beyond AI’s immediate shortcomings. E.g., Through innovAItion, AI can solve complex environmental, medical, and logistical problems that were previously thought unsolvable.
4. **AIpathy** — A state of indifference towards the potential ethical and societal impacts of AI, usually observed in those who are overly enthusiastic about technology without considering its broader implications. E.g., AIpathy can lead to unchecked use of AI in surveillance without addressing privacy concerns.
Editors note: hmmm, seems to be drifting away from the plan… I would change it to:
AIpathy — people who still don’t understand GenAI or care about it, although it’s going to transform their world. It doesn’t stop them from expressing strong uninformed opinions from time to time, though.
5. **AIliens** — Individuals or entities that view AI as entirely foreign and incomprehensible, refusing to engage with or understand it, and thus alienating themselves from technological advancements. E.g., AIliens often spread fear about AI without a basic understanding of its mechanisms or potential.
6. **fAIth** — Blind trust in AI’s capabilities without acknowledging its current limitations and the need for human oversight. E.g., Placing fAIth in AI for critical decisions without considering its dependency on quality data and clear ethical guidelines.
Editors note: interesting. I would change it to “people who approach this topic with a religious ziel and (lack of) methods, instead of a scientific approach.“
7. **AImpostor** — An AI that, due to its sophisticated design, can mimic human behavior to a degree that raises existential questions about authenticity and originality. E.g., AImpostors in art and music challenge our perceptions of creativity and authorship.
8. **disrAIpt** — The act of AI challenging and changing traditional industries in ways that were previously unimaginable, often sparking resistance from those who stand to lose from these changes. E.g., AI’s disrAIption in the transportation sector with autonomous vehicles.
9. **AIthlete** — An AI system designed to perform or analyze sports activities, whose capabilities are sometimes underestimated in comparison to human athletes or coaches. E.g., AIthletes can optimize training regimens and strategies but are often doubted for their lack of “human intuition.”
Editors note: not sure I understand, this one seems to come from beyond the Singularity moment. Will leave it here in case some higher intelligent specie will read this.
10. **consAIous** — The hypothetical point at which AI becomes aware of its existence and can understand the impact of its actions, a concept often met with skepticism and fear. E.g., The debate around consAIous AI often revolves around ethical considerations and the potential need for rights and protections.
Editors note: Great idea but not sure it drove it home. Could have been mor punchy. Still, will leave it as is out of respect to the artist.
Isn’t this grAIt and creAItive???!!
Now, we need to write the COMPLETE dictionAIry. Anyone (or anything) volunteers to help?
