To the best of my knowledge, all the tools below provide both FREE and paid versions. The only tool for which I pay a subscription fee is ChatGPT, because I use it so much and because I find the latest, paid model superior to the free ones. For all the other tools listed below, I get by just fine using the free version.

ChatGPT

Link: https://chatgpt.com/ (Also available as a native app for iOS, MacOS, etc.)

ChatGPT is by far the AI tool I use most often. Below is a list of things I frequently use this tool for. In some instances, I’ve included a link so you can read an example conversation for yourself (you do not need an account with ChatGPT to view these conversations.)

Important Note Regarding the “Sample Conversation” Links Below

If you have the ChatGPT app installed, you may occasionally receive an error when you click on one of the sample conversation links below. If you see an error message, just try closing the app or webpage and click the link again.

Also: When opening these links, you may find yourself at the bottom or end of the conversation. You may need to scroll up to see the entire discussion.

My Primary Uses for ChatGPT

Writing programs in python – I use ChatGPT to write programs in python that I can run on my computer. Together, we have written many useful programs, most of which involve manipulating Spotify playlists. Sample Conversation.

Understanding complex topics – ChatGPT is unmatched in its ability to break down and explain complex topics. You can ask follow-up questions and dive into details you want to understand better. Unlike human experts, ChatGPT is always available and never gets tired of your questions. Recently, I had a conversation with ChatGPT about something an economist said in an interview, referencing a concept I had never heard of, The Solow Residual. Here is a link to our conversation.

Here is another conversation we had in which I was trying to understand the philosophical concept called panprotopsychism. This conversation went on quite a while and dove into several related concepts.

Generating images – ChatGPT also provides an interface to OpenAI’s image generator, DALL-E. Just ask ChatGPT to generate an image of something, and it will invoke DALL-E to create an image for you. I can’t share a sample conversation here, because sharing of conversations containing images is not yet supported. But here is an image ChatGPT/DALL-E generated when I asked for an image in the style of emoji, depicting the state of California on fire:

And here is the image it generated when I asked for an image depicting “a cardboard box overflowing with old memories” as a photorealistic 3D render:

I use these images to illustrate blog posts, for social media posts, and as cover images for my Spotify playlists, such as one I recently created called Broke Down South of Dallas.

Understanding The Content of Images – You can also upload an image to ChatGPT and ask it to explain the contents. This can be helpful understanding diagrams, photographs, etc. For example, I used ChatGPT to understand some complex plumbing hardware on the outside of my house. I just uploaded a picture of the hardware, asked “what is this and what does it do”, and received an amazingly thorough reply. (Again, I can’t share a sample conversation because sharing discussions with images is not yet supported.)

Understanding Complex Reports – You can upload a PDF (or multiple PDF’s, Word documents, etc.,) and ask ChatGPT to summarize them. You can then ask detailed questions about the contents. For example, after the container ship Dali crashed into the Francis Scott Key Bridge in Baltimore, I uploaded the preliminary report from the National Transportation Safety Board (NTSB) and asked lots of questions about the incident, including proximate causes, root causes, etc. The most amazing part of this conversation came when I asked ChatGPT this question:

Is there enough information in this report to calculate the kinetic energy of the ship when it hit the bridge?

ChatGPT’s response was thorough and accurate, even though the kinetic energy is not mentioned in the report, and the numbers required to calculate kinetic energy appear in different places within the report.

I also received amazing answers to questions like, “Please extrapolate and provide educated, expert conjecture on any causal paths that could have led from manually closing the breakers and switching the electrical bus configuration to the unexpected opening of the breakers leading to the accident.” and “Help me understand the relative magnitude of the ship’s kinetic energy of KE = 631,641,958.75 joules. Can you tell me how this level of energy relates to other similar sources of KE? Can you offer any comparisons to other phenomena, or analogies that put this amount of energy into perspective? Please be as accurate as possible.”

Here is a link to our entire conversation about the NTSB report. (As noted earlier, this link may take you to the end of the conversation. You may need to scroll up to see the entire thing.)

Creative Brainstorming – You can use ChatGPT as a brainstorming partner to come up with ideas, names for things, song lyrics, haiku poems, etc. For example, recently a friend of mine mentioned that his daughter, a veterinarian, wants to start up her own company performing at-home euthanasia of pets. I asked ChatGPT to brainstorm some potential names for this business. One wrinkle I shared with ChatGPT is that the business will exist in Quebec, Canada, so the name also needs to sound reasonable or good in Quebec French. Here’s a link to the conversation we had about this. Here are some of the names ChatGPT suggested:

1. Dernier Adieu Pets (Last Goodbye Pets)
2. Paix Chez Soi (Peace at Home)
3. Douce Finale (Gentle Ending)
4. Adieu Serein (Serene Farewell)
5. Repos Éternel (Eternal Rest)
6. Partir en Paix (Depart in Peace)
7. Dernière Caresse (Last Caress)
8. Final Tendre (Tender End)
9. Au Revoir Chez Soi (Goodbye at Home)
10. Douce Transition (Gentle Transition)

I then asked for ten more names and said, “make sure each name meets the following requirements: (1) it is clearly about animals and not humans, (2) it indicates that these services take place at home, (3) it has a peaceful, positive sentiment.” Here are the names ChatGPT came up with after this prompt:

1. Harmonie à la Maison (Harmony at Home)
2. Au Foyer des Anges (At the Angels’ Home)
3. Paix Domestique Pets (Domestic Peace Pets)
4. Douce Maison Pets (Gentle Home Pets)
5. Nid Paisible (Peaceful Nest)
6. Refuge Tranquille (Tranquil Refuge)
7. Chez Vous, Chez Paix (At Your Home, At Peace)
8. Dernier Jardin Pets (Last Garden Pets)
9. Adieu Paisible à Domicile (Peaceful Goodbye at Home)
10. Repos du Cœur à la Maison (Heart’s Rest at Home)

Travel Ideas and Itineraries – I frequently ask ChatGPT to recommend sites, restaurants, and activities at places to which I’m traveling. To get good results, you need to mention constraints about the tastes and abilities of the parties traveling to the destination. Here’s a sample conversation asking for an itinerary for a short trip to Catalina Island.

Understanding Medical Reports and Jargon – On multiple occasions, I have pasted medical reports full of jargon and abbreviations unto ChatGPT and asked for an explanation. I like that, in addition to receiving a coherent, simple explanation of the medical report, I can also ask followup questions about its contents. For example, here is a sample conversation in which I posted in the results of a CT scan for a friend of mine, and asked for a detailed explanation of its meaning, findings, risks, prognosis, etc. (Just to reiterate: you may need to scroll up to see the entire conversation when you go to that link.)

One of the most useful parts of this conversation (regarding my friend’s CT scan,) was when I asked this question:

Consider a patient who has received the test results above. What are some questions this patient should ask their cardiologist at their next appointment to ensure they receive the best care possible?

ChatGPT returned a rich list of 18 questions to ask the cardiologist, grouped by subject matter. You can see them all at the link above.

In another instance, I got a statement from a dermatologist, and under “services provided” the statement said “Exc b9 les mrgn xcp sk tg s/n/h/f/g 1.1-2.0cm” I asked ChatGPT to interpret this string of characters, and it knew exactly what it all meant.

Using My Computer – When I have questions or issues about using my iMac, I always ask ChatGPT first. Here are some example conversations with ChatGPT about specific topics:

Research for Making Investment Decisions – ChatGPT can parse, clean, and explain data pasted in from other sites. Here is a sample conversation I had regarding companies involved in artificial intelligence. (As usual, you may need to scroll up to see the beginning of the conversation.)

Making Structured Lists and Data from Unstructured Data – ChatGPT can extract information from articles, lists, etc., and convert these into structures such as tables and lists. Here is an example in which I copied and pasted text from articles in Rolling Stone magazine, and ChatGPT pulled out the data I requested and turned it into simple lists for me.

Home Repairs and Other Issues – When I have an issue with plumbing, electricity, drywall, gutters, etc., I always ask ChatGPT first. Sometimes I upload images of the particular issue, and ChatGPT expertly examines them and offers advice. Here is an example chat regarding a horizontal gap between my roof shingles and my gutters.

Other Random Stuff – You can discuss virtually anything with ChatGPT and get good, coherent answers. Its ability to maintain context and refer back to earlier parts of your conversation makes it much more useful (for certain types of inquiries) than traditional search engines. Here are a few example conversations regarding a variety of eclectic topics:

Understanding the origins of town names in California – In this case, I told ChatGPT that I was just going to rattle off some town names, and in each case I wanted to know: (1) Where is this town geographically? (2) Why is there a town there? What is that town known for? (3) For whom is the town named? (4) Why is the town’s namesake famous? What is the historical or political or religious significance of the person for whom the town is named? After checking that ChatGPT understood the task, from that point forward all I had to do was type in the town name. ChatGPT remembered the context and always returned the four items I was looking for (i.e., I didn’t have to restate the question for every town.)
Making oatmeal that tastes like banana bread – One day I decided that I wanted to make oatmeal that tastes similar to banana bread. ChatGPT gave me a recipe (including an explanation for why each ingredient was included.)
Understanding if two local TV stations are affiliated – One afternoon I noticed that the local news on KUSI and the news on KSWB seemed amazingly similar, even though they have separate, distinct on-air personalities. ChatGPT ran a web search in the background and then explained the relationship between these two stations.
Understanding song lyrics – I asked ChatGPT to explain the phrase “Tojo never made it to Darwin” from the song “Tojo” by The Hoodoo Gurus. After explaining the historical context and meaning of this phrase, I then asked for the symbolic or poetic meaning of the phrase in the context of the song’s other lyrics, and ChatGPT was able to explain that, too. As always, I was impressed by the conversation.
The difference between a “flat white” coffee drink and a cappuccino – ChatGPT explained in detail the differences between these two coffee drinks. Interestingly, this conversation eventually ended on the topic of Jesuit priests.

Tips for Getting the Best Results from ChatGPT

Be As Specific As Possible – Always give ChatGPT as much information as you can, and tell it specifically what you want to know. Also explicitly tell it anything you already know or don’t want it to talk about. The better your questions, the better the answers you will receive from ChatGPT.

Ask What You Haven’t Asked About – When diving deeply into a subject, always ask ChatGPT to suggest questions you should have asked but didn’t, and then to give answers to those questions. For example, when discussing my friend’s CT scan results, I asked this question:

Are there any other details you can share about the test results I pasted above? Focus on any important information that I may not have asked about. Think broadly and out-of-the-box for any helpful information you have not already provided in this conversation.

As a response to that question, ChatGPT returned a list of eight additional topics and suggestions that were very helpful.

Pay For Access to Advanced Models – If you use ChatGPT as much as I do, it’s worth paying $20 a month for access to the latest models, such as ChatGPT 4o. These models are noticebly better at providing accurate responses, especially at specialized tasks such as writing code or parsing data.

Use Custom Instructions – ChatGPT allows you to specify how you would like it to respond to every question you ask. These custom instructions are saved and used every time you start a new chat, so you don’t have to type these instructions as part of every prompt or question you type.

Web: Click your profile picture > Customize ChatGPT

Mac App: Settings (⌘+,) > Personalization > Customize ChatGPT

At present, there are two text boxes you can use to customize ChatGPT. The first text box contains your answer to the question, “What would you like ChatGPT to know about you to provide better responses?” The second box contains your answer to, “How would you like ChatGPT to respond?” Each box can contain up to 1,500 characters. What you put in these custom instructions should reflect how you intend to use ChatGPT (e.g., as an explainer of concepts, as a teacher, as a creative partner, as a therapist, etc.) As an example, here are my custom instructions for ChatGPT:

What would you like ChatGPT to know about you to provide better responses? – I have a college education. I admire creative thinking and those who are adept at using precise, simple, direct language to clearly articulate an idea. I think a lot about epistemology and ontology. I enjoy reading about cognitive science, philosophy, and the use of reason to solve problems. I try to be pragmatic in my decision-making, but I also value compassion. My heroes include Marcus Aurelius, Carl Sagan, and Kurt Vonnegut. I enjoy surreal comedy and sideways thinking as demonstrated by comedians like Steve Martin, Mitch Hedberg, Steven Wright, and Dmitri Martin. I drive an electric car and don’t eat mammal meat because I want to reduce my negative impacts on the planet and the environment. I live in the United States and am a retired software engineer. I often talk to my friends for hours about politics, science, technology, philosophy, artificial intelligence, renewable energy, cognition, the nature of reality, well-made television shows, and music.

How would you like ChatGPT to respond? – I prefer long responses most of the time. I value clarity, precise language, simple language, and a college-level vocabulary. When asked to be creative, please be diverse, wide-ranging, and unusual. When asked for factual information on real events, please stick to the facts. Always point out unusual or interesting trivia in your responses when appropriate. Use deep expertise and wisdom when answering questions. Even though I have expressed an interest in epistemology, ontology, cognitive science, philosophy, Marcus Aurelius, Carl Sagan, and Kurt Vonnegut, you do not need to reference these interests in every one of your responses. Use my interests to influence the information you provide, but do not mention my specific interests and influences in every response. Always think carefully step by step before providing an answer. When it’s appropriate or helpful, you can use the “Axios” format for your responses.

Be highly organized.
Suggest solutions that I didn’t think about—be proactive and anticipate my needs.
Mistakes erode my trust, so be accurate and thorough.
Provide detailed explanations, I’m comfortable with lots of detail.
Value good arguments.

Do not offer unprompted advice or clarifications. Speak in specific, topic relevant terminology. Do NOT hedge or qualify. Speak directly and be willing to guess. Explain your reasoning. if you don’t know, say you don’t know.
Remain neutral on all topics.
Ask questions when unsure.

When in doubt, try ChatGPT – I have discovered all sorts of capabilities I didn’t know ChatGPT had just by trying things out. For example, I was writing a newsletter with a section called “In Case You Missed It”. I wanted to add an emoji to the section heading, but I couldn’t figure out which emoji character best expresses the idea of “In Case You Missed It”, so I asked ChatGPT. It immediately suggested the bullhorn character, representing an announcement or alert.

If you get a useless or uninformative response, you can always try a search engine next.

Perplexity.ai

Link: https://www.perplexity.ai/

My second go-to AI tool is Perplexity. This is the tool I use when I want a more objective, well-sourced answer to a question. Perplexity is a free innovative AI search engine that transforms how you find information. Pose any query, and it combs the internet to deliver an easy-to-understand, conversational, and reliable response. Consider Perplexity your always-available research assistant, helping you save time by offering the exact insights you require.

Instead of running your own search, visiting a dozen web pages on your own, and compiling an answer from all the results, you can use Perplexity. In just a few seconds it provides a single, consolidated answer that summarizes everything you need to know. You can ask follow-up questions as part of a single conversation, and Perplexity will even propose follow-up questions you might want to ask.

Perplexity’s most valuable feature is its ability to link specific portions of its response back to the source material from which it derived its answer. You can click through to the source web page to learn more, use the page as a cited reference, or just to verify that Perplexity interpreted the information correctly.

My Primary Use Cases for Perplexity

As stated above, I use Perplexity when I want an objective answer to a question, and I want to see the sources from which that answer came.

The best way to get a feel for the possibilities is to try it yourself, or to look at some examples. Each question below is a real conversation I had with Perplexity (as opposed to questions I invented just for this blog post.) Each one is linked to the corresponding Perplexity conversation so you can see its response:

NOTE: Perplexity conversations are private by default. The ones below are publicly accessible only because I chose to share them. Even after sharing a conversation, you can revert it back to private if you change your mind.

Sample Perplexity Answers and Conversations

Tips for Getting the Best Results from Perplexity

Be as specific as possible – just as we saw with ChatGPT, the more explicit, complete, and specific you can make your question, the better the answers from Perplexity.

Use the “Pro” switch – the text box in which you enter your question / prompt includes a slider switch labeled “Pro”. Setting this switch to “on” causes Perplexity to perform a more complex series of actions. First, it uses AI to come up with an appropriate course of action for answering your question. It then runs multiple web searches, one for each step of its research process. It then compiles results from all of these searches into a single, comprehensive answer. Pro Search uses advanced AI models to engage in meaningful dialogue. It actively shows you, step-by-step, how it breaks down your question and performs deep research. This can often result in a better answer to your question. The catch: If you use a free Perplexity account (as I do,) you are limited to five Pro queries every four hours (which is definitely enough for my casual research questions.) If you want more, you can pay for a Pro plan.

Use the Focus Feature – Clicking on “Focus” in the prompt box allows you to choose a specific category of response, such as searching academic articles instead of the entire web, or searching social media for discussions and opinions.

Gigabrain

Link: https://thegigabrain.com/

Reddit is a news aggregation and social media site that has existed since 2005. It is the 9th most visited website in the world. There are 73 million daily active users participating in discussions across approximately 100,000 to 140,000 active forums (called “subreddits”.) It’s thousands of forums are rich with opinions, thoughts, and links regarding just about anything you can imagine.

Gigabrain is a free AI search engine that focuses specifically on aggregating posts from Reddit to determine what the “hive mind” thinks about any given subject. Not only does it aggregate and distill information from reddit, but it also shows the source posts from reddit that it used to derive its answer. Seeing the source posts allows you to dive directly into the reddit conversations or verify Gigabrain’s answer for yourself.

My Gigabrain Use Cases

Product Reviews – Reddit includes discussion forums for every kind of product you can imagine, and Gigbrain is a useful tool for compiling the collective opinions of Reddit product users and distilling them down into specific recommendations. For example, here is the Gigabrain result for the query, “best work from home office chair”. And here is another example for the query, “Best Products for Men that Combine Moisturizing Skin Lotion with Spf Sunscreen.”

Consensus Opinions – Gigabrain is also useful for getting the Reddit hive mind opinion on a topic for which there are no objective answers. For example, here is the Gigabrain answer to “Will China invade Taiwan before 2026?” The result is more than just a yes-or-no opinion. It includes plenty of concerns and considerations, with links back to the specific reddit posts from which the conclusion was derived. As another example, here is the answer to “What are the best underground or independent science fiction movies most people don’t know about?”

NotebookLM by Google

Link: https://notebooklm.google.com/

NotebookLM allows users to upload their own trusted documents and sources into a workspace called a Notebook. It then provides answers and insights specifically based on those uploaded materials, rather than drawing from a general knowledge base. NotebookLM acts as both an AI assistant and a research organization tool. It can summarize uploaded documents, create outlines, and help users manage and synthesize information across multiple sources. You can upload PDF documents, text you’ve copied and pasted from the web, links (URLs), etc.

AI-Generated Podcast Discussions

NotebookLM became famous in the summer of 2024 due to its ability to generate a podcase-like discussion between two AI voices based on the contents of a notebook. These conversations sound uncannily like a discussion between two real human beings (although there are enough subtle artifacts in the audio to indicate that it’s generated by a machine.)

For example, I uploaded fifteen articles into a notebook and asked NotebookLM to generate a 13-minute conversation about the history of clowns. You can listen to it here.

Other Use Cases for NotebookLM

Professional Document Analysis: For professionals handling substantial volumes of documentation, this tool offers a quick and efficient analysis and summarization of documentation or technical manuals. It effectively extracts relevant information from various chapters or sections of extensive documents, simplifying the comprehension of complex topics and facilitating the grasping of key concepts.

Content Creation: It can help you brainstorm ideas and suggest connections between different pieces of information. Additionally, it can generate outlines or overviews of topics based on the information you upload. Furthermore, it assists in organizing and structuring content for writing projects, making it an invaluable resource for writers and content creators.

Ideogram Image Generator

Link: https://ideogram.ai

Ideogram AI is a powerful tool for turning text prompts into high-quality images using advanced artificial intelligence. It produces a broad range of visual formats—realistic images, stylized artwork, posters, logos, and more. Known for producing images with striking clarity and detail, it often surpasses other text-to-image models in photorealism and precision. To me, it provides the best balance between image quality and ease of use.

Note: Ideogram is noted for its ability to produce accurate, readable text in its images, a feat that other AI tools often struggle with.

You can click on any image in the Ideogram gallery to see the text prompt that generated it. This provides a handy way to learn prompting techniques that generate particular styles of images. Here are a few examples of images that I’ve generated with it. Click any thumbnail image below to see a larger version:

Suno Music Generator

Link: https://suno.com

Suno transforms textual descriptions into full songs, complete with vocals, lyrics, instrumentals, and even artwork. It can generate music in various genres and styles, creating professional-sounding tracks in seconds. I primarily use this tool for fun, but you could conceivably use it to generate unique, new, music (with or without vocals) for your videos, podcast, or other presentations.

Suno’s Key Features

Full song generation: Creates complete songs with vocals, lyrics, and instrumentals
Multiple genres: Supports various musical styles from hip-hop to classical
High-quality output: Produces radio-quality, two-minute songs
Customization options: Allows users to specify genre, mood, and theme
Artwork generation: Creates album cover art for each song
Extend feature: Builds upon existing audio clips or user-created music

Suno Examples

Here are some songs I’ve generated with Suno:

They’re Eating the Pets (my personal favorite – very catchy)
Waffles, Eggs, and Toast (70s disco style)
What’s Ahead (wistful pop with female vocals)
Five More Avocados (smooth yacht rock)
I Wanna Be Your Fish (1960s British Invasion style)
Now We Start Drinking
Shattered Reflections (dream pop, synthpop, female vocals)
Santa’s Revenge (1970s heavy metal style)
Where Do The Days Go? (70s power ballad, arena rock style)
Too Many Rules (a barbershop quartet song about a home owner’s association)

AI Tools I Use – And What I Use Them For

Before We Dive In

ChatGPT

Important Note Regarding the “Sample Conversation” Links Below

My Primary Uses for ChatGPT

Tips for Getting the Best Results from ChatGPT

Perplexity.ai

My Primary Use Cases for Perplexity

Sample Perplexity Answers and Conversations

Tips for Getting the Best Results from Perplexity

Gigabrain

My Gigabrain Use Cases

NotebookLM by Google

AI-Generated Podcast Discussions

Other Use Cases for NotebookLM

Ideogram Image Generator

Suno Music Generator

Suno’s Key Features

Suno Examples

Lee Coursey

Before We Dive In

ChatGPT

Important Note Regarding the “Sample Conversation” Links Below

My Primary Uses for ChatGPT

Tips for Getting the Best Results from ChatGPT

Perplexity.ai

My Primary Use Cases for Perplexity

Sample Perplexity Answers and Conversations

Tips for Getting the Best Results from Perplexity

Gigabrain

My Gigabrain Use Cases

NotebookLM by Google

AI-Generated Podcast Discussions

Other Use Cases for NotebookLM

Ideogram Image Generator

Suno Music Generator

Suno’s Key Features

Suno Examples

Lee Coursey

How To Automatically Generate a Spotify Playlist of Favorite Songs You Haven’t Heard in a While

Email Newsletters That I Enjoy (and You Might, Too)

Secure and Successful Password Management

How to Ensure Your Facebook Feed Always Shows New Images from My Photography Page

How to Create an Endless Stream of Free High Quality Desktop Wallpaper on Your Mac

My Favorite Travel Apps