by Tom Cloyd - 10 min. read - (reviewed 2025-02-23:1555 Pacific time)
Finding possible solutions to a problem is like hunting for something you are not even sure exists - you have to have faith and patience, but also some specific skills. The first part of the process, described here, focuses on understanding the problem of concern, and then on locating possible solutions. Do this first part well and you will much reduce your risk of settling for a poor solution when a better one may exist, or of not finding a solution at all.
With all serious problems, we are in the predicament of being in one condition but wanting to be in another. A good way to get at the specifics needed to address the problem well is to list what bothers you about your current situation.
The list you produce will likely have items that can be grouped in some sensible way. If this is possible it should be done. These groups may indicate fundamental aspects of your problem, or a group of several related problems.
With these groups, it is then possible to decide which is most critical and focus first on finding solutions for that one. This narrowed focus will make progress toward a solution much easier.
So, the first thing for you to do is to make this list of the basic aspects of the situation that is your problem. It is best to do this over at least two days. It is likely that not all aspects will come to mind immediately. While your list can certainly be amended at any point in the problem-solving process, it is most efficient to begin with a fairly complete list.
Make this initial list before going on to the next step.
Suppose you have a headache and go to a doctor for help, as it’s keeping you awake. He has two options:
Focus on your complaint (pain in your head), and give you medication to reduce that problem. OR…
Consider that this pain may be the result of something more serious which should not be ignored.
A well-trained doctor will choose option #2, for two reasons. First, treating only the symptoms of a more serious problem will at best provide only temporary relief. Second, ignoring the more serious problem may have grave results.
A wise physician will ask you questions about such things as recent falls or bumping your head against some hard object. If the headaches have been frequently occurring and increasing in intensity he may order a scan of your head to check for a possible tumor.
So, now that you see the difference between symptoms and their cause, look at your list of complaints and ask yourself: Is there something deeper going on here that should actually be the focus of my concern?
How you go about answering that question will be very dependent upon the context of your problem, but in many cases it may be wise to consult an expert. Their experience can make sense of your situation far more quickly and accurately than you can.
At this point, you are likely to make a common thinking error. I call this the “assumption of divinity” error. This happens when we basically assume that we know everything. This assumption is one we often make without realizing it.
Suppose I need a mechanic for my unusual imported racing bicycle. I think for a moment and conclude that there is no such person in my town, simply because I have never encountered such a skilled individual. But I cannot know this until I consult a directory of a all cycling shops in town, and then contact them to ask about their mechanics! I don’t have divine knowledge, and so have made a thinking error.
In this problem-solving procedure we will act to prevent this error by thoughtfully seeking out information sources beyond our own knowledge, as follows:
Why do this? Because your intimate familiarity with your problem may lead to especially creative solutions, in your imagination. You do NOT want to lose these ideas, and losing them is exactly what can easily happen when you start focusing on solutions from other sources.
Carefully list all solutions you can think of that have ANY real probability of success, even it seems small. If you’re uncertain whether to list something, be inclusive: list it.
AVOID being at all critical. Evaluation of possible solutions is a completely separate step in this process. Do NOT do that at this point!
If something works for someone else, you should consider using that something yourself, so it should be on your list.
At this point, you are still working alone, accessing your personal knowledge. Do not look at or consult outside sources, yet.
Add to your list any solution you know or have heard of that has worked for other people in your situation. Include any that almost worked. For you, it might work better.
This is a valuable thing to do with personal rather than objective problems for a surprising reason: psychologists have discovered that we only know about half of what can be known about ourselves. Other people see the rest! That means that they can add to our self-knowledge significantly, which can lead to changes in the possibilities we can see or accept for ourselves.
A “personal” problem might be that you are afraid of tests in school, or you have trouble getting a date.
An “objective” problem is that you car keeps needing repairs that you can’t afford, or your’re tired of dealing with the noise in your apartment complex.
With either type of problem you are wise to consult people you know who care about you. Explain your problem to them, then ask them for their ideas about how you can solve the problem. Make notes on what they say, thank them generously, then leave. Do not, at this point, evaluate what they tell you.
An expert is someone with extensive practical or theoretical knowledge of your sort of problem.
You do well to consult with experts simply because they have access to a much broader range of knowledge and ideas than you do. They are more likely than you are to see a solution to your problem.
How do you find such an expert? Ask people you know. Ask experts in other fields, because experts tend to respect expertise and know other experts. And ask the Internet - see the next step, below..
In the past 25 to 30 years, the Internet has made access to valuable information extraordinarily more easy for ordinary people, especially if they are in situations where access to good libraries or local experts is limited or non-existent.
Search engines such as Google or DuckDuckGo provide immediate access to an exceptionally wide range of information resources: magazines, newspapers, books, specialized databases, library catalogs, topical discussion forums, and social media.
I have written a special appendix to this article which details some valuable tips for doing quality Internet searches. Study it before beginning your searches, to take advantage of these tips.
There is a lot of discussion on the Internet about the new AI LLMs (Artificial Intelligence Large Language Models). In the last two parts of this section (sections 6.1 and 6.2), I offer two explanations of what they are and what you need to know to use them well, as a complete beginner.
From both reviews I have read and my own fairly active experience with these AI LLMs in recent months, I recommend consulting these AI models, in this order:
With each of these you will need to set up a free account. They also offer paid versions that are more powerful, but for now the free versions should be just fine.
My own personal favorites, based on the quality of the responses to my own questions, are Perplexity and Claude.
Remember, always, that AI model output should be considered a source for ideas, not a final say on anything. When a model’s output gives cited sources consider looking at them for verification and further information.
To get you started (and assuming you are a complete novice regarding AI LLMs), here is a question I gave to my two favorite models:
For someone entirely new to using AI LLMs, what are the essentials they need to understand to have a useful engagement with these models?
Their responses, in full, follows. I offer some comments about their output after that.
(I clicked the free “Pro” button for this response)
For someone new to using AI large language models (LLMs), there are several key essentials to understand for having a useful engagement with these models:
Understanding LLM Capabilities
LLMs are advanced artificial intelligence systems designed to understand and generate human-like text(1). They can perform a wide range of language-related tasks, including:
However, it’s important to recognize that while powerful, LLMs have limitations and potential biases that users should be aware of(2).
Effective Prompting Strategies
Be Clear and Specific
When interacting with an LLM, clarity is key. Be specific about what you’re asking or requesting to get the most relevant response(3). For example, instead of a vague query like “Tell me about planets”, you could ask “What are the diameters of the planets in our solar system?”
Provide Context
While LLMs don’t retain information from past conversations, they can understand context within a single interaction. Include relevant details in your prompts to help the model provide more accurate and tailored responses(3).
Specify Format and Detail Level
LLMs can adapt their output based on your preferences. Specify if you want a list, paragraph, or other format, and indicate whether you need a brief overview or an in-depth explanation(3).
Iterative Refinement
Don’t hesitate to ask follow-up questions or provide additional instructions if the initial response doesn’t fully meet your needs. Engaging in an iterative process can help refine the output and get you closer to the desired result(3).
Ethical Considerations
Be mindful of potential biases in LLM outputs and consider the ethical implications of using AI-generated content. Use the technology responsibly and critically evaluate the information provided(2).
Practical Applications
Understanding how LLMs can be applied in various fields can help you leverage their capabilities effectively. Some common applications include:
By keeping these essentials in mind, newcomers to AI LLMs can begin to engage with these models more effectively and harness their potential while being aware of their limitations.
Citations:
(1) https://hiddenlayer.com/research/the-beginners-guide-to-llms-and-generative-ai/
(2) https://geeksta.net/geeklog/prompting-ai-effectively/
(3) https://www.linkedin.com/pulse/unleashing-power-ai-6-tips-effective-communication-language-adams-1e
(4) https://www.msicertified.com/free-training/ai-prompting-essentials-certification/
(5) https://aws.amazon.com/what-is/large-language-model/
(6) https://www.moveworks.com/us/en/resources/blog/large-language-models-strengths-and-weaknesses
For someone new to using AI language models like myself, here are some key essentials to understand for having useful interactions:
We are AI, not humans: We don’t have human traits like emotions or personal experiences. We’re advanced language processing systems.
Our knowledge has limits: We have broad knowledge, but it’s not unlimited or always up-to-date. We may make mistakes.
We can’t learn or remember: Each conversation starts fresh. We don’t retain information from previous chats or learn new facts.
We can’t access external information: We can’t browse the internet, access databases, or open links.
Clear communication helps: The more specific and clear your questions or requests are, the better we can assist you.
We can handle various tasks: From answering questions to helping with writing, analysis, code, and more. Feel free to explore different uses.
Fact-checking is important: While we strive for accuracy, it’s wise to verify important information from authoritative sources.
We have ethical guidelines: We’re designed to be helpful while avoiding harmful or inappropriate content.
Follow-up questions are welcome: If our response isn’t clear or complete, feel free to ask for clarification or more details.
Experiment and explore: Try different approaches to see what works best for your needs.
Understanding these points will help you engage more effectively with AI language models and set appropriate expectations for our capabilities and limitations. Is there any particular aspect you’d like me to elaborate on?
Perplexity Pro (which is an option you can access with free Perplexity a limited number of times every day) produces lengthy, detailed output, with citations, as you can see. It has access to the Internet and so can provide very current information when needed. I routinely find its answers as good as any other models, and usually better.
Claude AI offers output that is terse, compact and well-focused, and augments that of Perplexity Pro nicely. This demonstrates the value of consulting with more than one AI model.
Using the Internet for research is a skill in itself. To get the best information, follow a methodical, structured search procedure. You are searching for the proverbial needle in the haystack, and you want to search as much of the haystack as possible.
Do this:
a. Write down and use search phrases that might get results. You want to try them all because you cannot know in advance which will be most profitable.
b. Use these phrases to search each of the information resource groups below.
These can usefully be grouped into those intended for general audiences (news magazines), limited audiences (men, women, military, sports, hobbies, etc.), and technical audiences (professions, research publications, etc.).
An example: In looking for ideas about how a poor single mother might improve her situation, I used this search string with the DuckDuckGo search engine: magazines for single mothers
I didn’t expect to get much, but I was very wrong! The results of search could easily take several days to consider.
These tend to focus on what is new and attention-getting. It is important to note that major newspapers publish for national and even international readers. At the same time, the national audience for a given country may be a fairly distinct groups of general readers.
An example: using this search string - newspapers women Kenya
Again, I didn’t expect to get much, and this time I didn’t. At the same time, there were indications of articles about women’s issues which could lead to names of experts, or people who know experts, so it would be wise to check these links and make notes on what is found.
One might expect not to have access to full-length books on the Internet, but such access definitely exists. The complete book is not always available, but one doesn’t need access to all of a book to benefit from a book’s contents.
Accessing books on the Internet can get a bit technical, but here are some starting points of serious value:
WorldCat.org - This remarkable resource accesses a database of major world libraries. This website is set up for everyone to use.
Use this resource to see if there are books on some topic, or to see where a particular book may be found.
With WorldCat, one selects a set of libraries that are local, or regional, or one can access very large libraries at major universities, just to see if something is available in some subject. As an example of what it covers, I can from my laptop see what books on a subject exists in any major city or university library in the USA (and other countries), but I can also access the book catalog of the libraries in the major local hospitals in my town.
Remember that is a book you want is not available in your local library, they may still be able to get it on loan from another library. This is a common practice in many countries.
Google Books - Read, locate, or buy books. An extraordinary resource!
This link provides access to digital scans of approximately 30% of the available book and magazine titles in the world (about 40 million titles). Their project started in about 2004 and is ongoing. Cooperating libraries, as well as publishers and authors, provide access to titles.1
If Google Books doesn’t have a scanned copy of a particular title, they will tell you what library near you has a copy. If there is no nearby copy, it offers a link to WorldCat (see description in previous section above) AND links to various online purchase options.,
Amazon Books - Originally an online bookstore, Amazon now sells almost everything, and has versions of itself for many countries.
They still sell books! More than that, many titles they sell allow you to look inside the book - but not at all the pages. Still, valuable information may be found using this tool. When you find such information, you will not be able to copy it but you CAN make a screenshot of it and save that image for later reading.
Once you are at the Amazon site, in the site header is a special search field for the site contents. It can help if you limit the search by selecting “Books” in the dropdown menu to the immediate left of the search field.
Locate a book title by entering the correct full title, part of the title, the general subject area, or the name of the author. Once you locate a specific book of interest, look directly to the left of the title itself for a picture of the book. Above that, look for the text “Look inside”, in orange and blue, with a small blue arrow (this option to look inside is not available for all books). Click that to open a window that allows you to look inside the book.
(coming)
Once we have a good problem description (our list of complaints), and a good collection of possible solutions, we are ready for Part 2 - Evaluating possible solutions, taking action, and evaluating outcomes, which is in development and will be available soon.
Wikipedia contributors. (2022, July 9). Google Books. In Wikipedia, The Free Encyclopedia. Retrieved 23:48, July 18, 2022, from https://en.wikipedia.org/w/index.php?title=Google_Books&oldid=1097167734 ^
☀ ☀ ☀