Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: How are you practically applying AI in your personal life everyday?
10 points by dominiconorton on Dec 28, 2023 | hide | past | favorite | 14 comments
Hi HN community,

With the rapid advancement and accessibility of artificial intelligence, I'm curious to learn about the unique and innovative ways individuals are integrating AI into their daily lives. Whether it's optimizing personal workflows, enhancing learning, managing smart home devices, or even delving into creative projects, the applications seem limitless.

Many of us hear about AI in grandiose terms – large-scale automation, complex data analysis, etc. However, I'm more interested in the practical, everyday applications that might not make headlines (beyond some of the AI hype).




I've tried to use ChatGPT (and Bard) for various day-to-day things, and honestly I have not found much of a place to fit them where they actually are helpful enough to actually incorporate them. I haven't tried the paid tier GPT-4 models, maybe those would actually be more useful, but the free tier has been less than impressive for me.

For writing tasks, I haven't been able to use it effectively for most tasks. It does fine on creative tasks but it doesn't really produce anything impressive enough to be middleware for writing. It's fine for writing emails and stuff though, just that I am not writing enough emails in a day for it to be a notable part of my workflow. I also don't like using it for technical writing. Expanding or rewording on technical content often leads to strange sentences, mostly because it tries to synonimize technical words that ideally shouldn't be tampered with in such a corpus... Maybe some better prompts would help with this, but I haven't had the chance to experiment more

I also tried it as a brainstorming tool, to try and pull out new ideas or to hold "discussions" so to speak. Again, it isn't bad at it, but often reverts to generic responses or occasionally repeats itself and it's just not efficient nor consistent enough for being a regular tool for this

The one thing that these models are exceptionally good at, however, is giving great overviews and explanations on concepts. It probably won't help you much if you are asking highly detailed, convoluted and specific things but they are amazing at "skimming" information. Like another user mentioned, it's good for obtaining domain knowledge on topics you're unfamiliar with. Kind of like a headstart for your learning process. I've found it very useful to get a general notion of "where to look" when I want to learn something new or run into hurdles on a project.


As an ideas generator. I often tell it to ask me 20 questions that will help me plan my week or something like that.

As the AI for the card game I'm developing. It knows the rules of the game so I can play test various scenarios really well.

As a travel planner. Queries like "Give me a 5 day itinerary for Luang Prabang, Laos" are a great way to get a broad overview of the main things to see in a new destination.

As a wordsmith. I like to invent fun new terms sometimes. For example: "Invent a word that's kind of like vegan but for people who only eat animals which would plausibly eat humans in nature." Response: "Predatovore".

As an emoji recommender. I don't want to sift through tons of emoji to find the appropriate one, so I just ask ChatGPT.


I have an emoji recommender here: https://github.com/smuzani/openai-samples/blob/main/node_exa...

It's wired into my command line. There's some crude instructions how to wire it into zsh on the README, but honestly, GPT-4 can give better.

Your other uses sound like they're common enough to use on the command line too.


I use it when I'm writing code.

Grandiose, large-scale automation has to start with a simple system that just works and is miles better than what it is replacing. To shrug off how AI changes the act of coding is, in my mind, to miss everything that matters about this technology.

When I use it, I'm in two halves. One part is engaged in the task itself (whatever I happen to be building). The other part is observing why the process feels comparatively so much better.


1. GPT-4 is excellent for step by step instructions. "How do I delete my account?" "How do I change the font in Y?"

2. Basically anything that needs large scale copy pasting. A website had a list of names, phone numbers, other IDs, and I just wanted to strip out the name into a list. If a spreadsheet doesn't work, then ChatGPT it is.

3. Brainstorming ideas. ChatGPT surprisingly sucks at this because it tries to be deterministic - ask it a question 10 times and it'll keep giving a similar result. The API is a lot better at it, especially completion models (davinci-002).

4. Basically as a "renderer" for https://random-character-generator.com/ The site determines whether a character is "brave", "strong", etc, and the AI figures out ways to "show, don't tell". It took me 7.5 hours to write 150 traits by hand. GPT-3 does 150 traits in 2.5 hours, together with quality checking and typo fixing. I can probably get it to write whole modules in GPT-4 today, but it's not a priority project now.


I wrote a post about this recently[0]. I took it slow with actually heavily adopting AI tools. Played around a lot and stayed up to date, but was in no hurry to pay for or rely on anything. But over time as the quality of available tooling began to meet my expectations, I've adopted following: GPT Plus, GH Copilot, Cursor Pro, and (intermittently) Elevenlabs.

[0] https://liza.io/how-generative-ai-has-fit-into-my-workflows/


I'm not using AI in my personal life everyday. I tried chatgpt several times as a help for writing. But I did not adopt it yet. I guess it wasn't helpful enough.


I'm in the opposite side! I find it extremely useful to improve emails and texts after writing the original draft.


I use it for writing simple but arcane code, e.g. things that involve the AWS SDK.

I sometimes ask it to improve my phrasing in English.

I use it to experiment with potential AI products.

I generate situational imagery and over-the-top poetry for comedic effect and share it with my wife.

I have it come up with weirdly detailed but ultimately useless logos for toy projects.

I translate text with it.

Sometimes I ask it about physics or biology, though usually just as a starting point.


I replace a class of search engine queries with chatgpt questions. I do this in particular when I dive into a new knowledge topic where I am a complete noob. That is where chatgpt shines for me right know, helping you navigate unknown territories.


I find it especially helpful when I don't know the topic enough to even know the terminology "what's that thing called where ..."

Another good one is completing lists "who are some artists similar to ..."


Literally gonna write the same thing. Instead of going through 10+ stackoverflow pages to fix linux bugs while still getting no explicit answer, I rather let BingGPT to figure it out for me in a second.


I think many times it has never been directly able to solve bugs for me but it has helped guide me on where to actually look. There's always those problems where sometimes you just don't know what to even write in the search bar, and ChatGPT can give you decent direction on the issue


"With the rapid advancement and accessibility of artificial intelligence, I'm curious to learn about the unique and innovative ways individuals are integrating AI into their daily lives. Whether it's optimizing personal workflows, enhancing learning, managing smart home devices, or even delving into creative projects, the applications seem limitless."

I feel like I'm developing a sixth sense for AI generated text - this particular paragraph has all the classic hallmarks of GPT.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: