On ChatGPT: "Doctor, Doctor"
Why am I Ragging on ChatGPT?
I feel like many of you know me well enough to understand that I often express my personal views in an “outside of the box” manner.
My opinions often indicate that I am forming a belief, building an understanding, and seeking a dialogue. Therefore, sometimes I may surprise you, even if you know me very well.
You might think, “wait a minute, I thought Nursey loved technology, of all kinds. Why is he ragging on ChatGPT these days?”.
I do love technology, of all kinds. I’m not ragging on ChatGPT, really, I am simply challenging the perception of intelligence, and the labels of wonder and amazement, which are attached to what is essentially a search engine. And it’s not even a very good search engine at that.
I sang to ChatGPT!
You don’t have to be a product of the British pop-music culture to see the irony here.
I sang to ChatGPT…
To which ChatGPT replied...
It is a very Aussie thing to say, “Oooh, you have tickets on yourself”. I mean, don’t you just want to say, “Call yourself an AI?”
Actually, the state of mind that leads to these posts may mean that seeking help from a medical professional right away, is actually sound advice. Fair call, ChatGPT!
Terrifying Video of AI playing drums
The hype and fear-mongering around AI is driving click-throughs like nothing else we’ve seen lately. AI taking jobs. AI helping students cheat in exams. AI writing gripping LinkedIn posts. Not this one, clearly. Oh before I forget, my favourite, "Terrifying video of AI playing drums". Terrifying? How do you think Kraftwerk shot to fame?
Truly, I wish I could be an exam marker who receives students work based on ChatGPT. Judging by the response to my singing, there would never be a dull moment!
Even though I am a geek, I don’t really care what the underlying technology of ChatGPT is. I only really care about what you can, or more to the point, shouldn’t do with it.
To be fair, as I’m not the AI-Police, you can do anything you want. Go cheat on your exam. Submit your next report to your boss, using ChatGPT generated text. Heck, apply for your next job with it.
领英推荐
AI Has to Know its Limitations
But you should know the risks and limitations.
Back to my point about not being a very good search engine, “Doctor Doctor can’t you see I’m burning burning.” ChatGPT, epic fail on that one.
But, Google got it right first time, and also credited the creators of this fiery pop song.
I’m yet to hear anybody say that use of ChatGPT led to a promotion, or winning a deal, or that it demonstrated any original thought. To wit:
...ChatGPT never fails to disappoint:
I think the request is very clear in indicating a desire to hear something original. And yet again, we have something plagiarised and not credited back to the human(s) that created it.
Fact Check Your AI
Just like a famous ex-US President, ChatGPT needs fact checking. If you’re heading into a chemistry exam and ask how to make invisible ink from copper sulphate, you’re probably going to get a decent answer. But if you’re behind a blast shield asking, “should I cut the red wire, or blue wire?”, well, you can see where I am going here.
One of the biggest legitimate fears about AI, is following one of its recommendations, and ending up in court, and not being able to demonstrate the process that led to the recommendation. i.e. a lack of auditability.
Another tangible fear is coded bias (look it up on NetFlix for an interesting watch). As an extension to that concept, if it’s not on the web, chances are that ChatGPT knows nothing about it. This leaves people who do not share knowledge through digital means, in an underrepresented, underserved position in society, a society which is beginning to believe all of our jobs are at risk, and all of our serviced needs will be met by AI and robots.
Like and Subscribe?
Will you become a paid subscriber to ChatGPT?
I for one won’t be bursting into the AI clinic, shouting, “Doctor Doctor!”
Be sure to comment below if you’ve asked ChatGPT anything, and then offered that as your own content. Have you been super-impressed, or mildly amused?
Which ChatGPT topics would you like to see explored in Part Two?
Airia - Enterprise AI Simplified
1 年You’re right about it not even being a very good search engine! I think a lot of the hype is due to folks ‘wanting’ it to be the next big thing. The expectation of time savings is coming at the expense of quality. I’ve asked it exactly two questions and it failed on both!
Data Governance | Director | GAICD | AGIA
1 年I'm glad it's not just me who has a healthy dose of skepticism. I'm reading a book about the myth of AI at the moment. I'm not a Luddite (cue: laughter) - I am a rational person who doesn't like overhyped things that will clearly deceive the masses and further reinforce the feeling that tech doesn't deliver on its promises. For round two, perhaps we could explore it's bardic capabilities?