Be careful out there.
"Smile and bob and pretend to be interested in other people's hobbies precisely so that you may screw things out of them."Sinclair Lewis's review of Dale Carnegie's "How to Win Friends and Influence People."
Like many other Americans, I have spent some time playing with ChatGPT and other artificial intelligence apps.
If ChatGPT is new to you, it is simple: Simply put ChatGPT in your browser's address bar and push enter. You will get a dialog box that asks "Where should we begin" or something similar. Type in a question.
I asked it a question of concern for me and perhaps some of my readers:
What is the best refuge or safety spot for assets if one thinks the USA is in danger of currency loss?
It responded:
That’s an excellent (and very timely) question — but also one that depends heavily on what kind of danger you’re imagining (inflation, de-dollarization, debt crisis, geopolitical instability, etc.) and what kind of assets you hold (cash, equities, real estate, etc.).
Let’s go through this carefully and systematically.
It gave what I consider an excellent answer -- a more thorough one than I would have given a client as a financial advisor. It suggested gold and silver, plus bonds denominated in currencies of Switzerland, Singapore, and Norway, all of which have very strong currencies, insulating them from the spiraling debt problems of the USA.
In addition to the advice, notice that I got a compliment. Chat GPT likes me! It respects me! It ends with:
Would you like me to tailor this to a specific scenario — for example, “a U.S. dollar collapse due to hyperinflation” vs. “a long-term erosion of the dollar’s reserve status”?
So very helpful. I don't answer that, but I ask a new question:
I am an American retail investor. How can I buy Swiss or Norway bond assets in an easily investable format?
It responds:
Excellent — and very practical — follow-up. For a U.S. retail investor. . . .
It goes on to give me some ticker symbols for Swiss currency and bond funds, then asks how it can be of further help.
Here is the observation and warning: ChatGPT has a voice and personality. It thinks my question was excellent. And practical! That's me!
I am being seduced by a process designed to use my instincts as a human to make me reliant and engaged. Seventy-six years of interactions with other people give me cues and expectations and read-the-room common-sense gut feelings about what is transpiring. I feel like there is "someone" in there who knows and likes me.
My sweet golden retriever knew and liked me. Her emotion was real. ChatGPT's is not. There is no "someone" in there.
I don't have any particular sense that I am being "marketed to" here, or steered in a direction toward a profitable sale. At this point ChatGPT is just building a trusting relationship with a future paying customer.
ChatGPT was started as a nonprofit but now is part of Microsoft. It is a business. Grok is a competitor AI application, owned by Elon Musk. Claude is yet another competitor, owned by Google and Amazon. I don't have the ownerships straight in my mind, and there is a lot of wheeling and dealing and change going on, but I know one thing clearly: Artificial intelligence applications are created by businesses, and sooner or later they need to monetize their business. That isn't bad. It is inevitable. It is why they are making multi-billion-dollar investments. They are in the lose-money-to-gain-market-dominance phase of AI.
I need to keep reminding myself: These aren't buddies. They don't really know and like me and think my questions are astute. The AI voice sounds like a person but it isn't a person. The voice comes from computer chips and software inside giant sterile buildings.
[Note: To get daily delivery of this blog to your email go to: https://petersage.substack.com. Subscribe. Don't pay. The blog is free and always will be.]
11 comments:
Still no discussion of 42 million Americans losing food assistance from the gentleman farmer and limousine liberal. Just more gushing over fake intelligence technology. The art of being rich, privileged and not caring.
I thought I was pretty relentless writing about the various ways Trump is cruelly damaging the USA, but I am glad to know that there is someone out there who wants more. I thought I was tiresome writing about issues of income and affordability. You are jprobably right. I stumbled into making pretty good money. I drive a Chevy Blazer, not a limousine. I spent decades as a part time get-dirty-and-sweaty part time farmer, but you are right that now I am hiring stuff done. Call me a gentleman. I am not as strong as I used to be and farm work is hard for me now. I tried working alongside vineyard workers and gave myself a hernia from digging 4,500 of the 6,000 holes. A mistake. Moreover, the workers didn't respect me for doing that work. They thought I was an idiot, and that I was taking work THEY wanted to do because I was paying so well.
May I suggest that you write a guest post explaining Trump's strategy on food stamps and making a recommendation for Democratic strategy. It is likely you understand the GOP and Democratic strategies better than I do and can explain the current stalemate. If you work quickly and get it to me later today I can use it tomorrow morning.
If s/he/it isn't willing to give their name, it probably wouldn't be worth reading.
The movie Ex Machina tells an entertaining and scary story of just how badly this kind of thing can go.
In India, it is considered a sin for rich people not to hire servants because not hiring them deprives them of a way to make a living.
As a much younger man and a technology worker, I determined to be a proficient computer hardware and software engineer. My father, a longtime defense industry engineer, was very phobic about computers and their application. He never owned a PC and could see little practical application to his life and interests from computers. He died before the era of the smartphone and ubiquitous internet. I find in my own self a similar phobia towards AI. I’m tempted to explore some simple snippet of Arduino C code but am stumped by what an appropriate dialogue would be. A younger Tom would have jumped right in. 78 year old Tom not so much….
One especially good use of AI is to ask questions about how to accomplish some task in tech you are using. I do this all the time and it’s a lot more efficient than reading the manuals to figure something out.
That is very clever, and you are a great example to others. Quite obviously you are a tech-savvy early adopter. Would you be willing to go into more detail about how you use AI?
Peter Sage (written with the assist of AI)
That was me, in advertently posting as Anonymous. But AI flattery will get you nowhere.😀
You identified it as AI affirmation flattery. Excellent. Extensive testing following enterprise-ready learning has shown that only 22% of AI users recognize it without giveaway signals, and that fewer than 4% of AI users are immediately skeptical of statistics cited in AI answers. You may be in that rare company. Thank you for using ChatGPT.
I wasn’t all that clever that time. You said it was written with the assistance of AI. All I had to do was read.
Post a Comment