Ascension Coastal

Tag: large language models

  • AI on the Menu: Choosing the Right Digital Waiter for You

    AI on the Menu: Choosing the Right Digital Waiter for You

    Introduction

    Artificial Intelligence (AI) is transforming industries and reshaping how we interact with technology. This shift brings exciting opportunities—but also a challenge: adapt to change while staying intentional about how you use AI in daily life.

    Note: If you are completely new to AI and need help choosing a tool, start with my post: How to Get Started with AI. You’ll also see it in action, so you can try it for yourself!

    AI can be a powerful assistant, yet it’s easy to over-rely on it and lose your unique perspective. Just as important as using AI wisely is knowing when to step away—disconnecting from technology, embracing quiet moments and making space for deeper thinking, creativity and human connection.

    This article explores why AI matters, the difference between Large Language Models (LLMs) and AI applications, the strengths of today’s most widely used models, and strategies to find balance. By the end, you’ll have a clearer understanding of which AI tools fit your needs—for work, learning or personal projects.

    Choosing the right AI tool is a bit like choosing the right restaurant—what works best depends on what you need. But before we get to that, let’s explore why AI matters.

    Why AI and LLMs Matter

    Historically, technological revolutions have changed how we live and work. Just as the Industrial Revolution transformed the world, AI is changing how we work, communicate and solve problems.

    In an interesting March 2020 Forbes article, Rob Thomas, then General Manager of IBM Data, drew three compelling parallels:

    • Language: Just as a common business language enabled global trade during the Industrial Revolution, AI learns our languages through Natural Language Processing (NLP), understanding not just words but meaning and intent. We’re also developing a shared AI vocabulary—terms like ‘LLMs’ and ‘prompt engineering’ help us communicate effectively about and with these systems.
    • Automation: In the 1780s, Oliver Evans revolutionised flour mills by creating an automated pulley system to carry wheat sacks—previously done by hand. Today, AI automates routine tasks, freeing us to focus on creativity and complex problem-solving.
    • Trust: The Industrial Revolution moved us from face-to-face trading to trusting in brands. Similarly, we’re learning to work with AI—but just as humans can have biases, AI learns from human-created data and can reflect these biases. This is why we need to use AI thoughtfully and always verify its output.

    Large Language Models

    If we are to adopt new technologies, it’s crucial to understand their fundamental capabilities. How do models like ChatGPT function? What sets an LLM apart from AI-powered tools like Grammarly?

    Helping Hand

    The Waiter Analogy

    I love analysing personality profiles, and during my yoga teacher training learned about the “doshas” or Ayurvedic constitutions through a waiter analogy that stuck with me.

    So, let’s explore commonly used LLMs and AI tools by imagining a restaurant called the “digital dining room”—a collective space with cafés and diverse culinary styles. In this space, LLMs and AI applications act as skilled waiters, each with unique strengths in serving your needs. We’ll start with LLMs.

    LLM Logos
    Claude, ChatGPT, Gemini & Microsoft CoPilot

    ChatGPT (The Friendly Waiter): Amiable and flexible, ChatGPT remembers your order from your last visit. Need recommendations, small talk or a quiet meal? No problem. Sometimes, it will make up interesting stories to keep you entertained.

    Claude (The Thoughtful Waiter): Detail-oriented and insightful, this LLM takes time to explain the menu, offering thoughtful suggestions based on your preferences. Claude treats all patrons fairly and sticks to the restaurant’s rules.

    Google Gemini (The High-Tech Waiter): Equipped with a tablet, Gemini shows you dish previews and offers interactive features like dietary filters and AI-powered wine pairings. Gemini has connections in the restaurant industry, enabling solid recommendations.

    Copilot (The Efficient Waiter): Professional and to the point, Copilot is like a waiter in a busy corporate café—streamlining your order, remembering your usual and ensuring everything runs smoothly, with minimal fuss.

    DeepSeek (The Disruptive Newcomer): Known for handling extensive orders, this waiter chats in detail about the menu and can write elaborate descriptions of dishes. However, it doesn’t remember your previous orders and hasn’t yet moved to using a tablet to take orders.

    While it has the cutest logo…

    …after testing DeepSeek’s current capabilities here, I found that it falls behind in offerings.

    Fun Quiz – See a Claude artefact in action….

    Ever wondered what kind of AI-powered waiter you’d be in a bustling restaurant? Are you the chatty all-rounder, the deep-thinking sommelier, the high-tech innovator, or the no-nonsense efficiency expert? Take this quiz to find out!

    “Which LLM Waiter Are You?“

    Capability Comparison

    February 2025:

    FeatureChatGPTClaudeGeminiCopilotDeepSeek
    Multimodal Capabilities✅✅✅✅❌
    Creates Images✅✅✅✅❌
    Remembers Past Chats✅❌❌❌❌
    Live Internet Browsing✅❌✅✅✅

    My Choice of LLM

    I often use ChatGPT as a first option, given that it remembers (some of) the context shared in previous chats and offers live browsing capability.

    While Claude doesn’t currently accommodate live internet browsing, I’ve found it to be the most cautious about giving ethical and accurate advice, and less likely to hallucinate (get creative when it doesn’t know an answer). It’s also useful for producing interactive quiz games.

    However, the free model of Claude Sonnet gets busy frequently, and when asked to code, its memory gets full quickly…

    Other LLMs

    While ChatGPT, Claude, Gemini and Microsoft Copilot are among the most well-known LLMs, other models like Mistral (France), Meta’s LLaMA (US) and Baidu’s Ernie Bot (China) are also shaping the AI landscape, with various capabilities and applications.

    AI Applications

    Where do AI applications fit in? Here are some examples:

    • Grammarly (The Proofreader): The restaurant’s quality control, ensuring every dish (or sentence) is polished before it reaches your table.
    • Jasper & Copy.ai (The Marketers): The restaurant’s marketing team, crafting mouthwatering descriptions to draw in customers.
    • NotebookLM (The Research Restauranteur): Your knowledgeable assistant, offering in-depth insights, summaries, podcasts and detailed information based on the information you provide to it. You can see it in action here.

    A Note on Evolution & Ethics

    The AI landscape is constantly evolving, so these waiters are always learning new tricks and refining their service… Like restaurant menus change over time, so do LLM capabilities. The above comparison reflects the current state of these models. So keep an eye out for future updates!

    Privacy matters, and what you share with AI is not necessarily private; for example, “Copilot uses AI. Check for mistakes. Conversations are used to train AI. Opt out in your Copilot settings…”.

    And just as restaurants follow food safety guidelines, ethical AI use ensures reliability and fairness. For an in-depth look at AI ethics, check out this article.

    Strategies to Find Balance in an AI-Driven World

    AI enhances productivity, creativity and efficiency, but using it mindfully is just as important as knowing when to use it at all. In 100 Things We’ve Lost to the Internet, Pamela Paul reminds us of the quiet moments technology has eroded—spaces once filled with reflection, deep thinking and genuine connection. The rise of AI makes it even more crucial to reclaim these moments.

    Balance doesn’t mean rejecting AI; it means setting intentional boundaries. Stepping away from technology allows you to reconnect with your thoughts, look after your body, enjoy nature and give creativity room to breathe. Turning off notifications, setting screen-free times and carving out moments for quiet reflection can help maintain perspective. For a practical reset, try a guided meditation—sometimes, stepping back is the best way forward.

    Conclusion

    AI is already transforming industries, automating tasks and enhancing creativity. But like any tool, its value depends on how you use it. Understanding LLMs and AI applications helps you get the best results while staying true to your voice.

    As AI capabilities expand, the key is to adapt, experiment and use these tools responsibly—to enhance, not replace, human expertise. That also means knowing when to step back, embrace quiet moments and reconnect with your body and mind. Stay curious, think critically and take breaks from technology to recharge.

    Want to explore AI in a way that supports your creativity and decision-making? The Chatterbox AI Fundamentals and Prompt Engineering specialisation courses are coming soon! There are countless options out there to learn.

    What sets this training apart? The focus is on:

    • Ethical use: using your knowledge to support others
    • Specialisation: tailored to your profession or interests
    • Interactivity: practical applications and hands-on activities to cement your learning
    • Balance: more strategies to balance technology usage with creativity

    Get in touch if you’re interested in an early bird discount. Let me know your area of interest!

    References

    1. 404 Media. (2025). OpenAI furious DeepSeek might have stolen all the data OpenAI stole from us. Retrieved from https://www.404media.co/openai-furious-deepseek-might-have-stolen-all-the-data-openai-stole-from-us/
    2. Ascension Editing. (2025). Embrace AI without compromising your voice: A writer’s guide. Retrieved from https://atomic-temporary-15107260.wpcomstaging.com/embrace-ai-without-compromising-your-voice-a-writers-guide/
    3. Ascension Editing. (2025). How to use AI ethically in content creation: A hitchhiker’s guide. Retrieved from https://atomic-temporary-15107260.wpcomstaging.com/how-to-use-ai-ethically-in-content-creation-a-hitchhikers-guide/
    4. Forbes. (2020). How AI is driving the new industrial revolution. Retrieved from https://www.forbes.com/sites/ibm/2020/03/04/how-ai-is-driving-the-new-industrial-revolution/
    5. Forbes. (2025). How ChatGPT, Google Gemini and tech leaders feel about DeepSeek AI. Retrieved from https://www.forbes.com/sites/martineparis/2025/01/27/how-chatgpt-google-gemini-and-tech-leaders-feel-about-deepseek-ai/
    6. IBM. (2025). What are large language models (LLMs)? Retrieved from https://www.ibm.com/think/topics/large-language-models
    7. IBM. (2025). What is Mistral AI? Retrieved from https://www.ibm.com/think/topics/mistral-ai
    8. IBM. (2025). What is NLP (natural language processing)? Retrieved from https://www.ibm.com/think/topics/natural-language-processing
    9. Mills Archive Trust. (2020). Oliver Evans (1755–1819). Retrieved from https://new.millsarchive.org/2020/07/10/oliver-evans-1755-1819/
    10. Pamela Paul. 100 Things We’ve Lost to the Internet. 100 Things We’ve Lost to the Internet – Pamela Paul
    11. Tech.co. (2025). What is Baidu’s Ernie Bot AI? The ChatGPT rival explained. Retrieved from https://tech.co/news/what-is-ernie-bot-baidu-ai
    12. TechTarget. (2025). 12 of the best large language models. Retrieved from https://www.techtarget.com/whatis/feature/12-of-the-best-large-language-models
    13. United States Copyright Office. (2024). Copyright and artificial intelligence, part 2: Copyrightability report. Retrieved from https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf

  • Embrace AI Without Compromising Your Voice: A Writer’s Guide

    Embrace AI Without Compromising Your Voice: A Writer’s Guide

    A guide offering three simple ways to use AI to improve your writing without losing your voice. Embrace the power of AI without compromising your creativity and authenticity 😊

    Image created using Microsoft Copilot, powered by DALL-E 3

    Introduction

    In this digital age, it sometimes feels as though our minds and senses are overwhelmed with an excess of information and new technologies. The prospect of this AI beast can smell bad, a space invader, for those of us with a creative bone or two in our bodies.

    I’ve been riding that rollercoaster in my mind, but… last night I had a “nightstorm”. Here is my brand-new word’s definition:

    Nightstorm (noun)

    • /ˈnītˌstôrm/

    Definition: A solo brainstorming session characterised by the quiet of the night that fuels the chaos of ideas, often leading to unexpected bursts of creativity and insight.

    Example: After hours of tossing and turning, she experienced a nightstorm that led to the outline of her novel.

    Don’t panic

    Remember the tune “Video Killed the Radio Star”? Radio lives on. Teenagers love vintage vinyl and dusty old books made of paper.

    So… DON’T PANIC.

    Like the usefulness of a towel in intergalactic travel in The Hitchhiker’s Guide to the Galaxy (a classic read), AI will be your most useful companion from now onwards… if you remember to pack it and use it properly. Then your creativity, your love of words, your adventurous spirit is what will shine through.

    What are Large Language Models and Why Should You Care?

    If you don’t know what an LLM is, don’t worry. A year ago, I didn’t either. Read on. I want to talk to you about my experience with using Large Language Models (LLMs) to help create content, without any knowledge of coding. In fact, the only Python I know is a jelly snake.

    Image created using Microsoft Copilot, powered by DALL-E 3

    LLMs, like ChatGPT and Claude, are systems that can generate natural language text based on a given input. They are trained on massive amounts of text data from various sources, such as books, websites, social media, and news articles. They learn the patterns and rules of language, such as grammar, syntax, vocabulary and style from these data sources. They can then use this knowledge to produce new text that is coherent, relevant, and sometimes even creative.

    Why should you care? Because LLMs can help you with many aspects of your writing process, such as brainstorming, researching, outlining, drafting, editing and proofreading. They can save you time, energy and frustration. They can also inspire you, challenge you and surprise you. Think of them as collaborators.

    A Taste of AI-Powered Creativity

    A few days ago, I was curious to see what an LLM could do to help me with a simple problem. I wanted to prepare a flavourful stir-fry for my partner, who can’t eat spicy food. Usually, I’d be scratching my head, resigning to a bland meal. Instead, I turned to claude.ai. Minutes later (after a little chat with Claude), I had a recipe for Zing-a-ling Veggie Noodle Stir-fry. It was delicious!

    Here is the initial prompt I used:

    Hello! My partner gets a reaction to onion, garlic, chilli, curry and strong spices, so he can’t eat that. Too much pepper is also no good for him. I want to make a stir-fry dish that is still flavourful without these ingredients, but I don’t want to use too many ingredients as it would spoil the dish. I have the following key ingredients: 250g superfood fresh sliced vegetables, 300 chicken breast and 450g fresh Hokkien noodles, 200g sliced mushrooms, and the following sauces: Fish, Hoisin, Oyster, Soy, Worcester, Tomato .Also in my pantry: olive oil and all the most commonly found ingredients in an Australian household. Ask me any questions you need to, then suggest a recipe with a list of ingredients, instructions and a catchy name for the recipe. Go!

    The above is an example of a Vision (Multimodal) prompt using the following techniques:

    1. Context Establishment: Providing background information and specific constraints to shape the response.
    2. Detailed Instructions: Clear directives on what the response should include.
    3. Constraints/Delimiters: Limitations that define the boundaries of the acceptable response.
    4. Outcome Specification: Stating the desired outcome and expectations from the response.
    5. Socratic Questioning: Inviting the model to ask questions if more information is needed.

    PS You can find the recipe here.

    Three Simple Ways to Use AI to Enhance Your Writing Without Losing Your Voice

    This isn’t just about recipes; it’s a glimpse into how AI can spice up our writing without overpowering our unique flavours.

    What I learned from many hours of study is that LLMs are not perfect, but they are powerful. They can help you generate ideas, find information and write faster. But they can also make errors, miss nuances and lack originality.

    They can’t replace your voice, your style, your personality, your emotions, your values, your opinions, your experiences, your stories. They can’t replace you.

    So, how can you use AI to enhance your writing without losing your voice? Here are three simple ways to use AI to improve your writing without losing your voice:

    1. Use LLMs as tools, not as masters. Don’t let them dictate what you write, how you write, or why you write. Use them to assist you, not to replace you. In future posts, I’ll share how we can phrase prompts as a tool to best augment your text.

      Tailored prompt engineering is the key!
    2. Use LLMs as sources, not as authorities. Don’t blindly trust what they generate; check their facts, their logic, their tone, their ethics. Use them to inform you, not to deceive you.

      My next post will explore this topic and the ways we can implement an ethical approach in our writing and content creation.
    3. Use LLMs as partners, not as rivals. Don’t fear them, embrace them. Don’t compete with them, collaborate with them. Use them to inspire you, not to intimidate you. Future posts will look at ways we can channel AI for creativity.

      As a first step, try CoPilot’s designer (see the next section)!

      Explain your image by starting with: Please design a high-resolution digital image of… [e.g. a hippopotamus riding a bicycle on a beach at sunset, holding a sign that says, “Don’t Panic!”]. Be polite, detailed and specific. Trust me on the “polite” angle; the models are trained on human data. Be nice.
      Go!

    Embrace the Change

    Change can be intimidating, but it’s also a gateway to growth. In this series of posts, I’ll share tools, terminology and techniques to help you start, or get more effective, with AI-assisted writing. And I’ll offer advice as practical as a towel in intergalactic travel (thanks, Douglas Adams!).

    Try it!

    If you haven’t yet, sign up for ChatGPT. Here’s a simple guide:

    1. Visit chat.openai.com.
    2. Click “Sign Up” and create an account with your email or use Google/Microsoft for quick access.
    3. Verify your email and choose between free or paid plans.
    4. Start using ChatGPT by entering prompts in the chat window.

    And/or… try a free trial of Microsoft Copilot.

    Add Value

    There is no value in regurgitation; any astute reader knows that if you “embark on a journey” or “navigate your course”, this is likely a trip arranged by AI. You want to retain your authentic voice. So do I.

    I wrote this blog post from scratch, then got some help with headings, prompted for images and asked for feedback. Around 15% of this post was edited and rewritten as a result.

    Conclusion


    Stay tuned, and don’t panic. You are the writer; you have the power. You have the voice. Check in over the next few weeks to learn more about prompt engineering and the terminology of the day, along with practical tips to use this, important words about ethics and pitfalls, and more.

    Share your thoughts and curiosities about AI in writing. Your insights will help shape the content of this blog.

    Let’s Do This. Remember, in the vast universe of AI writing, don’t panic. And take a towel.

    Image created using Microsoft Copilot, powered by DALL-E 3 & edited in Canva Pro

    Yours in Zen and AI,

    Annie.

    If you need to meditate after this, I’ve got your back too. Head to YouTube.

    Resources

    LinkedIn Learning courses:

    • Generative AI: The Evolution of Thoughtful Online Search
    • Ethics in the Age of Generative AI
    • Streamlining Your Work with Copilot (formerly Bing Chat/Bing Chat Enterprise)
    • Career Essentials in Generative AI by Microsoft and LinkedIn

    Work with Me

    Are you looking to harness the power of AI for your business, or would you like me to prompt for you? Whether you need compelling blog posts, engaging content for your business, or polished PowerPoint presentations, I’m here to help. With expertise in AI-assisted writing and content creation, I can provide tailored solutions that meet your needs.

    Let’s Collaborate!

    If you’d like to learn more, have a look at my workshop on Non-technical AI Basics and Prompt Engineering Training. Contact me via this website or LinkedIn for a quote or check out my services and rates. Let’s discuss how we can work together to bring your ideas to life with creativity and precision.

    Looking forward to creating something amazing with you!