# What practical tasks are AIs already good at today?

**URL:** <https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737>\
**Category:** In My Humble Opinion\
**Tags:** ai\
**Created:** [March 24, 2026, 8:49pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737 "2026-03-24T20:49:09Z")\
**Posts on this page:** 20\
**Page:** 1

<div class="post-metadata">

**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:** [March 24, 2026, 8:49pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/1 "2026-03-24T20:49:09Z")

</div>

## Background

This is an attempt to create a ongoing thread that documents the latest advances in consumer AIs (LLMs or otherwise).

By consumer AI, I mean tools that are free or relatively affordable (\< $100-ish a month, let’s say) and which are meant for the general public, not hidden behind enterprise contracts or confidential military secrets.

This is not meant to be a philosophical debate about the pros and cons of AI (we have plenty of those threads already), just it’s practical everyday uses.

In other words, **what useful things have you discovered an AI could do for you lately**?

## Things AIs are good at - March 2026

For example, just to start the thread…

- We already know they can [create cool images](https://boards.straightdope.com/t/ai-image-generation-is-getting-crazy-good/1017203)
- They are very good at translating between commonly-used human languages (any LLM, really — it’s like a basic baseline function for them)
- They’re also decent at making music now ([Suno](https://suno.com/), [Gemini](https://gemini.google/overview/music-generation/))
- They can make detailed SVG animations ([Gemini 3.1 Pro](https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/)) like for this [optical illusion](https://boards.straightdope.com/t/optical-illusion-or-clickbait/1028309/9)
- They can [help track airline ticket prices and email you a daily summary](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/113)
- They can [split PDFs](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/119)
- They can [help with text file format conversions](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/112)
- They’re **really** good at computer programming, especially if you use coding “agents” like [Claude Code](https://claude.com/product/claude-code), [Google Antigravity](https://antigravity.google/), or [OpenAI Codex](https://openai.com/codex/) (instead of just a chatbot)
- They can [control your web browser directly if you ask it to](https://developer.chrome.com/blog/chrome-devtools-mcp)
- They are pretty decent at making human-sounding speech and copying voices (e.g. [Elevenlabs](https://elevenlabs.io/)) or understanding speech ([Whisper](https://openai.com/index/whisper/), [Parakeet](https://developer.nvidia.com/blog/turbocharge-asr-accuracy-and-speed-with-nvidia-nemo-parakeet-tdt/)). Some of those tools are also OK, but not perfect, at dubbing videos to other languages while keeping the person’s voice and accent.
- They are very good at summarizing scholarly or scientific sources and producing readable summaries, podcasts, Powerpoints, and video explanations ([NotebookLM](https://notebooklm.google.com/))
- They’re OK but not great at making short videos ([Veo](https://deepmind.google/models/veo/), [Sora](https://sora.chatgpt.com/), [Kling](https://klingai.com/global/))

What else?

---

<div class="post-metadata">

**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:** [March 24, 2026, 8:54pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/2 "2026-03-24T20:54:33Z")

</div>

To be honest, the coding I described in the linked post was done via chatbot (Gemini 3.0, Pro).

> [@So... How does one actually use AI?](https://boards.straightdope.com/t/so-how-does-one-actually-use-ai/1027842/118):
>
> As this thread is more appropriate, I’m going to quote something I posted about in the Pit AI thread regarding a couple of more use cases for which I found AI to be very helpful. As it’s long, I’m going to put this in a details tag. play_button I’ve taken the bottom example and now have a suite of web apps where we can upload documents, create a journal entry, and then put that JE into QuickBooks online. Processes that once took an hour, hour and a half, now take 3, 4 minutes.

---

<div class="post-metadata">

**Author:** ![Procrustus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/procrustus/32/2994_2.png) [@Procrustus](https://boards.straightdope.com/u/Procrustus)\
**Post date:** [March 24, 2026, 8:55pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/3 "2026-03-24T20:55:12Z")

</div>

I just used co pilot to answer some questions about 600 pages of medical records. I got answers in about 15 seconds.

---

<div class="post-metadata">

**Author:** ![Richard\_Pearse](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/richard_pearse/32/13144_2.png) [@Richard\_Pearse](https://boards.straightdope.com/u/Richard_Pearse)\
**Post date:** [March 24, 2026, 9:02pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/4 "2026-03-24T21:02:33Z")

</div>

The trick then is to verify them.

I’ve been using a new budgeting tool that doesn’t yet have the ability to export a small subset of transactions as a .csv file or similar. I’ve found ChatGPT to be useful for turning a screenshot of the transactions into an excel file. I can also just upload it to the previous chat where I told it how to format the first one and tell it do the same again rather than explaining everything each time.

---

<div class="post-metadata">

**Author:** ![Procrustus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/procrustus/32/2994_2.png) [@Procrustus](https://boards.straightdope.com/u/Procrustus)\
**Post date:** [March 24, 2026, 9:12pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/5 "2026-03-24T21:12:09Z")

</div>

> [@Richard\_Pearse](#):
>
> The trick then is to verify them.

Absolutely. One of the answers surprised me and so I asked copilot to give me the page numbers and I then could very quickly verify (it was correct).

---

<div class="post-metadata">

**Author:** ![Dereknocue67](https://avatars.discourse-cdn.com/v4/letter/d/6de8d8/32.png) [@Dereknocue67](https://boards.straightdope.com/u/Dereknocue67)\
**Post date:** [March 25, 2026, 8:18am UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/6 "2026-03-25T08:18:48Z")

</div>

I had great success with AI as I went to small claims court and won a judgment against a contractor who had damaged my property. It walked me through the forms necessary to file, with suggestions on how to briefly submit my case to the judge, produce documentation, and specific steps to pursue in order to collect and garnish wages. I recovered my damages and court costs.

---

<div class="post-metadata">

**Author:** ![wolfpup](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/wolfpup/32/10618_2.png) [@wolfpup](https://boards.straightdope.com/u/wolfpup)\
**Post date:** [March 25, 2026, 10:31am UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/7 "2026-03-25T10:31:57Z")

</div>

I’m going to go out on a limb here and go against the grain of some popular opinions expressed here on the Dope, and say that an AI like ChatGPT is genuinely useful as a general-purpose information resource. Yes, it can be wrong, yes, it can hallucinate and bullshit at times, but in my experience those events are sufficiently rare and ease of verification sufficiently easy that the net utility of ChatGPT as an information resource is very high.

I’ve had it interesting discussions with GPT ranging from cosmic issues about the universe to resolving household issues about appliances and cooking, all of which were very useful and informative.

I was impressed by recent interactions about my new (well, new to me) car. The owner’s manual is huge and not particularly well organized, and GPT was able to answer questions about how certain things worked which I know were correct because things worked as it said they did.

I was particularly impressed when I asked it about the many different adjustments and controls on the power driver’s seat, which the manual described only with a diagram with a bunch of arrows which weren’t particularly obvious. When I asked GPT, it responded much like a human would – “show me the diagram”. Since I also had a PDF of the owner’s manual, I was able to cut the diagram and upload it, whereupon GPT explained exactly what all the controls did, including the insight that one of them was a multi-directional control that I should conceptualize as a joystick rather than a lever.

---

<div class="post-metadata">

**Author:** ![HMS\_Irruncible](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/hms_irruncible/32/7394_2.png) [@HMS\_Irruncible](https://boards.straightdope.com/u/HMS_Irruncible)\
**Post date:** [March 25, 2026, 1:02pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/8 "2026-03-25T13:02:38Z")

</div>

I have mainly found it useful for:

- Generating ideas as raw material for analysis
- Adding useful context to other texts
- Spotting errors and bullshit in other texts

This last bit may be counterintuitive, because while we all know that LLM’s are copious generators of the most ridiculous bullshit you’ve ever seen, it’s quite capable of spotting flaws, issues, and logical inconsistency in texts that it didn’t originate.

Which is more or less expected if you think about it… LLMs would be incredibly slow and wasteful if they were built to second-guess everything they said. Recently some LLM tools have been modified to expose a little of their internal deliberations by way of showing progress, and if you direct them to be skeptical, you can easily see them spiraling out into prolonged internal conflict and second-guessing. It’s got to be cut off in the name of efficiency, and the results often aren’t satisfactory.

However if you provide it with a piece of text and say “find the bullshit”, or “find the contractual loopholes” it will easily surface a list of things that should be checked more closely, and is usually correct about the flaws it finds (or close to it).

So while AI is a dangerous trap for the credulous, it’s a pretty powerful tool for the skeptically-minded.

---

<div class="post-metadata">

**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\
**Post date:** [March 25, 2026, 1:29pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/9 "2026-03-25T13:29:42Z")

</div>

Never mind. Out of scope for this thread, as opposed to the other AI threads.

---

<div class="post-metadata">

**Author:** ![enipla](https://avatars.discourse-cdn.com/v4/letter/e/54ee81/32.png) [@enipla](https://boards.straightdope.com/u/enipla)\
**Post date:** [March 25, 2026, 1:41pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/10 "2026-03-25T13:41:55Z")

</div>

Well, I sent a long time friend a picture. She asked what kind of tree it was in a response email.

The AI suggested response in Gmail was wrong. It tried to format an email to a very dear friend I’ve know for decades.

I was stunned. Took me a minute to figure it out, but I got it turned off.

---

<div class="post-metadata">

**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:** [March 25, 2026, 2:06pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/11 "2026-03-25T14:06:49Z")

</div>

> [@Dereknocue67](#):
>
> I had great success with AI as I went to small claims court and won a judgment against a contractor who had damaged my property.

That’s pretty cool. Was it able to mostly “one-shot” it, or did you have to have a prolonged back-and-forth? How much detail did you have to give it?

> [@HMS\_Irruncible](#):
>
> So while AI is a dangerous trap for the credulous, it’s a pretty powerful tool for the skeptically-minded.

I had an interesting experience with this last night, trying to annotate a list of cars with their engine types. I pitted AI against AI in doing so, first starting with a coding agent (which wrote a Python script to look each car model up in a .CSV file I fed it), then feeding the output of that through a regular LLM which caught some parsing/English errors the coding agent did not, like some contradictions between the engine and notes columns.

Its ability to detect, even if just probabilistically, “this does not follow”-type situations is genuinely impressive and fascinating to me. I wish I could see into their latent space and better understand their reasoning… the math (and theory) of it all is above my intellect.

> [@LSLGuy](#):
>
> Out of scope for this thread, as opposed to the other AI threads.

I’d love to hear whatever you wanted to say, if you do post it elsewhere.

> [@enipla](#):
>
> She asked what kind of tree it was in a response email.

Yeah, I’m still skeptical of this approach. Even before LLMs, image-based tree identification wasn’t very reliable, especially if you didn’t have closeups of the different parts.

Some trees look very distinct while others can be easily mistaken at a casual glance. The more science-y apps usually ask you to take individual photos of the trunk, branches, leaves, fruits, etc., and maybe count or describe certain parts of them to double-check.

I suspect the general-purpose image models don’t have sufficient specialized training for that and probably just end up correlating tree pictures with random social media posts, which is maybe good enough for some trees in casual conversation, but not a very good scientific approach to tree ID.

---

<div class="post-metadata">

**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:** [March 25, 2026, 2:09pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/12 "2026-03-25T14:09:30Z")

</div>

> [@Procrustus](#):
>
> I just used co pilot to answer some questions about 600 pages of medical records. I got answers in about 15 seconds.

How did you feed it 600 pages of medical records? Was it able to keep all of that in its context window just from a regular file upload, or did you have to use specialized tooling?

---

<div class="post-metadata">

**Author:** ![enipla](https://avatars.discourse-cdn.com/v4/letter/e/54ee81/32.png) [@enipla](https://boards.straightdope.com/u/enipla)\
**Post date:** [March 25, 2026, 2:25pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/13 "2026-03-25T14:25:44Z")

</div>

> [@Reply](#):
>
> I suspect the general-purpose image models don’t have sufficient specialized training for that and probably just end up correlating tree pictures with random social media posts, which is maybe good enough for some trees in casual conversation, but not a very good scientific approach to tree ID.

It just pissed me off that it DARED to suggest a response to a long time friend. It’s totally out of its wheel house there.

It wasn’t just that the tree was wrong.

It suggested what I should write as well. @#$% that.

---

<div class="post-metadata">

**Author:** ![Dr\_Paprika](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/dr_paprika/32/3042_2.png) [@Dr\_Paprika](https://boards.straightdope.com/u/Dr_Paprika)\
**Post date:** [March 25, 2026, 2:28pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/14 "2026-03-25T14:28:51Z")

</div>

I don’t fold a lot of proteins, but AI is said to be helpful to pharmaceutical companies for making accurate 3D models of complex things, similar molecules, and evaluating potential drugs for market.

Is it hype? We will see if new antibiotics are soon coming, to replace the ones with significant antibiotic resistance.

---

<div class="post-metadata">

**Author:** ![Procrustus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/procrustus/32/2994_2.png) [@Procrustus](https://boards.straightdope.com/u/Procrustus)\
**Post date:** [March 25, 2026, 2:41pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/15 "2026-03-25T14:41:25Z")

</div>

> [@Reply](#):
>
> How did you feed it 600 pages of medical records? Was it able to keep all of that in its context window just from a regular file upload, or did you have to use specialized tooling?

I got 10 years of a child’s records in one PDF. I just dropped it into the box. (I first had to confirm that this was a “private” AI platform)

---

<div class="post-metadata">

**Author:** ![Ponderoid](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ponderoid/32/19221_2.png) [@Ponderoid](https://boards.straightdope.com/u/Ponderoid)\
**Post date:** [March 25, 2026, 2:54pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/16 "2026-03-25T14:54:00Z")

</div>

It still might have overwhelmed its context window. Whenever context gets long, I occasionally double-check by asking it specific questions about various parts of the context. And it no longer surprises me how often it gets things wrong, even when by strict word/token count, everything is supposedly still within its context window.

---

<div class="post-metadata">

**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:** [March 25, 2026, 3:33pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/17 "2026-03-25T15:33:26Z")

</div>

> [@enipla](#):
>
> It suggested what I should write as well. @#$% that.

Oh yeah. You can turn off those Gmail smart features if you don’t want them, especially “Smart Compose”: [How to opt out and turn off all ‘smart’ AI features in Gmail](https://www.idownloadblog.com/2026/01/30/turn-off-smart-ai-features-gmail/)

I don’t like that either (the forced AI)… all the companies are just shoving it down everyone’s throats. I much prefer to use them on my terms, only when I want to.

---

<div class="post-metadata">

**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:** [March 25, 2026, 3:36pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/18 "2026-03-25T15:36:24Z")

</div>

> [@Procrustus](#):
>
> I got 10 years of a child’s records in one PDF.

> [@Ponderoid](#):
>
> It still might have overwhelmed its context window.

Yeah, I’m with Ponderoid here. It’s not clear whether it was actually able to read that entire PDF… you might want to ask it “what’s on page XXX (near the end)”, for example, along with a few pages in the middle, to make sure it actually read that far. (They usually don’t/can’t, but they don’t warn you about it beforehand.)

I try this often with SDMB threads, for examples, and it only reads the first handful of posts and never reaches the end of longer threads, despite being _absolutely confident that it had reached the last post_. (They don’t know what they don’t know, and they don’t have a good way to self-evaluate their level of confidence.)

---

<div class="post-metadata">

**Author:** ![Ike\_Witt](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/ike_witt/32/283_2.png) [@Ike\_Witt](https://boards.straightdope.com/u/Ike_Witt)\
**Post date:** [March 25, 2026, 3:56pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/19 "2026-03-25T15:56:27Z")

</div>

Based on all the practice they have been exposed to, I imagine that AI should be really good at detecting pictures of traffic lights.

---

<div class="post-metadata">

**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\
**Post date:** [March 25, 2026, 3:59pm UTC](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737/20 "2026-03-25T15:59:41Z")

</div>

> [@Reply](#):
>
> I try this often with SDMB threads, for examples, and it only reads the first handful of posts and never reaches the end of longer threads, despite being _absolutely confident that it had reached the last post_. (They don’t know what they don’t know, and they don’t have a good way to self-evaluate their level of confidence.)

I wonder how much that specific issue has to do with the on-demand scrolling feature of the Discourse html UI? Doubly so if the AI is scraping the human UI rather than requesting the thread from whatever API Discourse exposes.

We’ve gone around in Site Feedback a time or two about how to obtain the entire thread body in one go regardless of post count. The consensus is that at least within the browser UI there’s no way to do that.

[Next page](https://boards.straightdope.com/t/what-practical-tasks-are-ais-already-good-at-today/1028737.md?page=2)