# So... How does one actually use AI? **URL:** \ **Category:** In My Humble Opinion\ **Tags:** ai\ **Created:** [February 20, 2026, 3:06am UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842 "2026-02-20T03:06:13Z")\ **Posts on this page:** 11\ **Page:** 19
**Author:** ![JohnT](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/johnt/32/15048_2.png) [@JohnT](https://boards.straightdope.com/u/JohnT)\ **Post date:** [October 23, 2026, 4:06pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/361 "2026-10-23T16:06:14Z")
Taking a page from @Reply’s post above, I uploaded the sample ballot for our address and had Claude explain the propositions. Sometimes these things are worded so that you don’t know what they’re for, and/or you don’t know if your “yes” vote approves or disapproves the initiative. ---
**Author:** ![Reply](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/reply/32/15952_2.png) [@Reply](https://boards.straightdope.com/u/Reply)\ **Post date:** [October 23, 2026, 5:53pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/362 "2026-10-23T17:53:38Z")
It also helps to have it trace funding sources. Many measures are from dark money trying to fool with you with deceitful wording and unclear downstream impacts… some states have an independent analysis, but even that only goes so far. ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [October 24, 2026, 1:46pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/363 "2026-10-24T13:46:10Z")
Another thing I’ve been using it for lately is a scheduled task checking every day for specific art calls, concerts coming into town, and jobs I might be interested in (I’m a self-employed photographer). I give it parameters, plus it knows me well enough by know to know what opportunities I’m interested in, and which don’t reach my threshold for payment, visibility, reputation, etc. It found one gallery (not a pay-to-play one) a couple months ago and one of my photographs was accepted and is currently being exhibited there in a group show. I also have about ten irons in the fire with outstanding submissions to various galleries I would not have known about or would have taken much too much time finding and curating were I googling on my own every day. It’s quite handy. Yesterday, for an art call for the main branch of our library here in Chicago, it helped me put together a formal 20-page PDF proposal (most pages were photos; about 4 were text). Not in the sense that I just had it do it for me, but it helped me organize it. I had it go page by page and tell me what to write or design for that page, and then I would write the text myself, asking for help where a word or phrase was on the tip of my tongue. So essentially using it to give me an outline while I fill in the blanks. And then I had it double check my work (it found several inconsistencies in my captions, as well as a spot where I accidentally gave two different works the same title), and with a quick human proofread and sanity check by me, it was sent off, two or three hours later. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [October 24, 2026, 1:55pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/364 "2026-10-24T13:55:13Z")
When you ask it to prompt you with new information, how does that arrive? Does the app/browser show a notification, do you get an email or a text, etc.…? ---
**Author:** ![pulykamell](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pulykamell/32/3166_2.png) [@pulykamell](https://boards.straightdope.com/u/pulykamell)\ **Post date:** [October 24, 2026, 2:03pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/365 "2026-10-24T14:03:39Z")
It sends me a push notification and I also have it set up to send me an email. Like here’s the email I got ten minutes ago: " Chicago Photo Leads Chicago Photography Lead — October 24, 2026 New direct-client opportunity: March 2027 wedding — $3,500 budget A couple posted Friday, October 23, looking for a wedding photographer whose strengths include candid coverage, difficult light…" This it found through Reddit and the local wedding pages. I passed on this one, as the while the candids and difficult light are two of my strengths, client wants a film/film-like aesthetic, so I also prompted it to reject any calls for vintage-style photography or anything that relies on heavy post-processing look. This particular prompt is less for aggressive client acquisition but more just to see if there are any crumbs out there I could sweep up. These are just bonus jobs. ---
**Author:** ![Maserschmidt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/maserschmidt/32/18829_2.png) [@Maserschmidt](https://boards.straightdope.com/u/Maserschmidt)\ **Post date:** [October 24, 2026, 2:04pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/366 "2026-10-24T14:04:26Z")
Thanks for the explanation! ---
**Author:** ![Crafter\_Man](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/crafter_man/32/458_2.png) [@Crafter\_Man](https://boards.straightdope.com/u/Crafter_Man)\ **Post date:** [October 24, 2026, 2:57pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/367 "2026-10-24T14:57:11Z")
> So… How does one actually use AI? Certainly not to _write_ something, IMO. At work, I have read some reports from coworkers where it’s instantly obvious they used AI to “spruce up” something they’ve summarized or written. It’s laughable and horrible at the same time. I must admit, however, that the AI summary resulting from a Google search has been occasionally helpful. ---
**Author:** ![pjd](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/pjd/32/3009_2.png) [@pjd](https://boards.straightdope.com/u/pjd)\ **Post date:** [October 25, 2026, 10:26am UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/368 "2026-10-25T10:26:23Z")
I have recently been asking Google AI to explain how to perform specific tasks in Studio One recording software. It provides simple step-by-step click-by-click instructions which have made progress a lot faster than it would have been by me trawling through manuals or YT videos. ---
**Author:** ![ratatoskK](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ratatoskk/32/2987_2.png) [@ratatoskK](https://boards.straightdope.com/u/ratatoskK)\ **Post date:** [October 25, 2026, 4:04pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/369 "2026-10-25T16:04:53Z")
The NY Times just tested AI agents Muse (from Meta) and “dots” (from OpenAI / ChatGPT). They used for to make restaurant reservations, apply for jobs, plan meals, and haggle for power prices. Dots costs a lot and Muse is free and worked better. [Here is the article](https://www.nytimes.com/2026/10/25/technology/personaltech/openai-dots-meta-muse-4-killer-tasks.html?unlocked_article_code=1.LVE.ShFf.ERn1yk12L-Qs&smid=url-share). ---
**Author:** ![Reply](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/reply/32/15952_2.png) [@Reply](https://boards.straightdope.com/u/Reply)\ **Post date:** [October 25, 2026, 4:55pm UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/370 "2026-10-25T16:55:05Z")
Heh, sounds like what Google tried to do with the Google Assistant a few years back (before LLMs), where their creepy voice bots would call restaurants to try to make bookings for you. Guess we’ve moved right past the uncanny vanny into the “yes, m’lord” phase of AI acceptance 😆 From the article: > So why not ask a bot to sit at a computer and wait for the tables to open up? Well, because then eventually it just leads to bot scalpers who buy up all the slots as soon as they appear, for everything from [campsites](https://www.kqed.org/news/11450483/cant-get-that-camping-spot-it-could-be-bots) to [concert tickets](https://www.yahoo.com/entertainment/music/articles/scalpers-vs-fans-concert-tickets-205932078.html). One bot releases it, another books it, another resells it, another bids for it at a secondary auction… heh, the endgame is just turning daily life into high-frequency trading. ---
**Author:** ![billy-jack](https://avatars.discourse-cdn.com/v4/letter/b/ac91a4/32.png) [@billy-jack](https://boards.straightdope.com/u/billy-jack)\ **Post date:** [October 26, 2026, 11:22am UTC](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/371 "2026-10-26T11:22:49Z")
Using AI to generate your program code can be a big issue. From [Chinese AI Writes Weaker Code for Beijing’s Adversaries. We Fixed It. A Qwen3-Coder Case Study](https://www.hirundo.io/blog/qwen3-coder-vulnerabilities) > **TL;DR: Chinese AI models now sit inside Western products, and some carry geopolitical triggers: tell the model who the code is for, and the security of what it writes changes. For the first time, we showed that this vulnerability lives in the model’s weights, and that removing the model’s CCP alignment removes it.** … > Security researchers had already found early evidence of backdoors in these models. CrowdStrike and Booz Allen each reported Chinese coding models writing more vulnerable code when a prompt mentioned a topic or a customer Beijing opposes. There’s a discussion about this at: > **[Bias in Qwen3-Coder Model - Schneier on Security](https://www.schneier.com/blog/archives/2026/10/bias-in-qwen3-coder-model.html)** > > Interesting research: TL;DR: Chinese AI models now sit inside Western products, and some carry geopolitical triggers: tell the model who the code is for, and the security of what it writes changes. For the first time, we showed that this... [Previous page](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842.md?page=18)