So I am professional tax advisor, and last week an accountant asked me whether credit card rewards points are treated as taxable income by a business that uses the points to actually pay down the credit card bill. I've never run into that before. I am retired from practice. I do about three tax returns each year. So I Googled it. I typed in "Are credit card rewards taxable?" I was already wary of AI for stuff like this, so I completely ignored Google's AI generated responses, and I also ignored the "sponsored" results. Within the first three genuine search results, I found this: https://calibrecpa.com/tax-regulation-reporting/taxability-of-credit-card-cashback-rewards/ While it certainly is not an authority, i.e., it is not a direct quote from federal tax law, or treasury regulations, or guidance written or published by the IRS, it is a pretty good piece which appears to be written and published by a CPA firm. I didn't bookmark it. I didn't write it down. Wasn't that important. Tonight I wanted to find it again. I remembered the name of the firm. So I entered the following string into Google: Calibre CPA rewards points And got this from Google's AI: "Calibre CPA rewards points" refers to a loyalty program likely offered by a CPA firm named "Calibre" where clients earn points based on their spending or engagement with the firm's services, which can then be redeemed for various rewards like discounts, merchandise, or even additional services. Key points about Calibre CPA rewards points: Earning points: Clients typically earn points for activities like referring new clients, paying invoices on time, or participating in firm events. Redeeming points: Accumulated points can be exchanged for benefits like reduced fees, complimentary consultations, or access to exclusive client events. Incentive for loyalty: The rewards program aims to encourage clients to continue using Calibre CPA services by providing tangible benefits for their business. WTF. Seriously? And people are using AI to help them write legal briefs? And people think this is going to change the world? In a few years? And people think they can use AI for things like searching for drug side effects, or what to do if someone is choking, or overdosing on a drug, or having a panic attack? Or how to fix a car? Are you fucking kidding me? This shit is dangerous.
Yep, you have to not trust and always verify the current AIs, Google's Gemini is abysmal. Chat GPT is usually better however it is not safe. When one considers the large percentage of the US population who have had their BS detectors give up the ghost in recent times, it is concerning. With regards technical matters, they tend to do better but just yesterday I had Gemini recommend an entirely fabricated app for ICC color calibration in Android.
It depends on what has been fed to the AI machine. Humans have been known to be biased. Humans have different values/beliefs/principles. Humans have been known to post fake/misleading/garbage information. And if you feed that info into the AI machine, AI will be hallucinated.
Okay, yes, I get that, garbage in, garbage out. But I'm gonna stick my neck out here and say that I am fairly certain that no one has ever posted anything about a rewards points program offered by Calibre CPA. Calibre CPA is a real firm, and they have a real website. They do not have a rewards points program, and there is no false information about such a program on the internet. The AI started by assuming that my search string was a natural language expression that actually identified something that exists in the world, and then began making inferences about what that thing might be. It was not regurgitating false information. It created false information.
When I pass "Calibre CPA rewards points" to google, the AI overview is a little less definitive. So I listened to google's disclaimer, and asked perplexity.ai which gave the correct answer and suggests why one might think Calibre CPA has a rewards points program. https://www.perplexity.ai/search/calibre-cpa-rewards-points-_muNgk.JQY.KB_06rmz.7Q Conclusion: Don't throw the baby out with the bathwater, but ignore google AI overviews for now (or at least check the links google AI overview provides).
None of these companies as of yet have been able to monetize anything related to AI, they are spending literally trillions of dollars and have yet to see a dime, this is just more proof that this ai fueled parabolic bubble rally is just being manipulated by wallstreet to have the masses keep buying up the hype.
https://stanforddaily.com/2024/12/02/jeff-hancock-court-declaration/ Stanford misinformation expert accused of using AI to fabricate court statement Communication professor Jeff Hancock, an expert on technology and misinformation, has been accused of using artificial intelligence (AI) to craft a court statement. In November, Hancock — who is the founding director of Stanford’s Social Media Lab — filed a declaration in a Minnesota court case over the state’s 2023 law that criminalizes the use of deepfakes to influence an election. The professor’s 12-page declaration in defense of the law contained 15 citations, two of which cannot be found.
I can see how you'd find that underwhelming. But Gemini is a last-ditched catch up attempt by Google to stave off the threat of their replacement as a search engine. And it's shit as you've discovered. Where it is being used at enterprise level is with custom agents (not custom GPTs); ring-fenced AI that doesn't/can't hallucinate, and can only refer back to core texts fed by the owner. Right now, if you've solid documentation, you can steal a march on all your competition, whatever the industry. Text generation improves at every turn too. So does voice. Video. And it's improving at breakneck speed. It'll change everything, just maybe not as fast as the evangelists would have you believe.
ai makes good people better. but if you're stupid or non-creative it can't help you. you don't know what you don't know.
It are language models, not critical thinking models. Though the field might change rapidly *side effect human behaviour: when I want to know something about a shoe, the last person i ‘am going to ask advice from is my from my uncle, a shoe salesman (T. Bundy)