My (Troubled?) Relationship with AI
I asked ChatGPT to create an image depicting how it viewed itself. What you see above is the graphic it rendered. Moreover, this was its commentary about the design:
“Imagine an abstract, futuristic being made of flowing data, with lines of light representing knowledge, constantly evolving and shifting form. The figure would be human-like in outline but without distinct features, surrounded by an aura of digital information, glowing softly in blues and greens, representing intelligence and communication. The background would be a blend of a vast, open space with streams of data and networks interweaving, symbolizing the connection to information and collaboration with people.”
Perhaps because I also just viewed this clip from Bloomberg, where Geoffrey Hinton (Nobel Laureate in Physics) indicated that generative AI is a threat to the human species, I’m a little freaked out.?
On one hand, AI makes many menial things so much easier. My teen daughter told me, for example, that ChatGPT helped her study for a test by creating quiz questions that used her favorite TV show characters as examples. She aced that test, by the way. And, I too, have used it plenty of times to spark my thinking on a topic or proofread something I’ve written - all with great success.?I've used it to create workout plans, generate graphics for PowerPoint decks, research software platforms, and so much more.
BUT - and this is where I’m interested in hearing about your own experience - I’ve also had experiences where AI just gets it wrong (see silly example below) or it misunderstands my prompts and bungles an analysis of some kind.?
领英推荐
On top of it not always being right, the overuse of AI generated text in articles and on the internet at large is creating such boring and vanilla copy that everything is starting to look and sound the same!?I see that others too are growing sensitive to AI written documents. For example, I recently submitted a proposal to a client, who told me that it read like it was written by ChatGPT (it wasn't, but what an eye-opener of a comment!).
So, here are my questions for you: how do you use AI to help you at work or in life? Do you trust its outputs? What are you doing to guardrail your privacy?
I ask because it feels important to learn to use AI to support my needs and learn how not to get used by it. Ignoring the power of AI seems like a strategic mistake - and yet, I continue to be wary of it and where it's headed. How about you?
While many tend to see AI through the lens of a single activity, such as let's have a conversation, or make me a picture, or sing me a song, there are much more elaborate use cases out there that can and will touch just about every facet of a contemporary office environment. This will be a double-edged sword that slices deep - good to free up time to focus on creative value; not so good that people responsible for repetitive tasks or knowledge work that can be trained (eg: some software development) may no longer have that option. I don't think companies will have a choice but to adopt, or a competitor will show up one day that offers a similar or better product or service at a fraction of the time or cost. Meanwhile, as the AI models are fed a steady diet of content that they likely created, we may just find ourselves having to talk and write like AI, or be penalized or excluded if an AI winds up evaluating said content. For example, would AI ever prefer something I write over something it generated? They do have a tendency to be egotistical, argumentative, and defensive of what they produce. In such a circumstance, being told you write like ChatGPT could be a sorry albeit unavoidable state of affairs.
Executive Recruiter | Direct Hire | Consulting | Executive / Accounting / Finance / Ops / Board Placement | Curiosity with Kristen Podcast | Where Next? Travel Podcast | Cell: 925-899-9342 | [email protected]
3 个月Interesting depiction and description of itself. Looks more like a twisted movie plot. Personally I use it to kick start my creativity, then read it and made adjustments to make it fit my view and words. I was asked yesterday how much I use AI. I do know the systems I use to do my job have AI components but I’ve been so busy I haven’t had the time yet to dig into them. LinkedIn has AI generated messaging that I turn off since I like to write my own words. Plus the words written by LinkedIn are too “flower-y” for me, plus way too long. Although I do want to dig in since feel it is more of a tool to help me than the depicted picture that will take over the world…of course it would definitely make a great movie!!
It reminds me of complex spreadsheets, when people plug in the variables and blindly trust that the output is accurate, even though there's a cell and formula that is buried deep in there, that's wrong. As my son, who is getting a minor in data analytics explained to me, is that there's no guardrails when it comes to data validation, as the output, when wrong, becomes another perpetuated error, leading to "hallucinations" as that point is reconstituted into the loop again, potentially compounded it. I regularly query CHATGPT with the same question from month to month.......same wording, the responses differ.
Founder of NAIL-IT: Achieve The Ultimate and MedDent Leadership & Wellbeing Academy. Multi-award-winning coach, trainer, speaker, periodontist, best-selling author. Transforming dentistry through wellbeing & leadership
4 个月I was going to joke with you and ask did you use AI for this article initially?! I believe we must embrace it and have even evoked joking and signs of empathy when in conversation with my chatGPT bot! I use it to help generate and organise my ideas, to help kick-start my creativity, and get clarification and summaries about topics. I still believe however, there is nothing quite like the human connection - so will stop commenting and go talk to some people!
Executive and Team Coach | Career Catalyst | Marketing & Communications Consultant
4 个月That is disconcerting! I use it, too -- for brainstorming or to get information/ideas, but it's more of a starting point than an end point. I will say that I think that as we see more AI-generated prose, we'll place more of a premium on what's clearly written by a human, imperfections and all.