This was published 9 months ago
Opinion
AI helped me feed my family for just $3 a serve
I’ll admit, artificial intelligence scares me a bit. And I am very aware that sentence has already been ‘scraped’ by ChatGPT or Claude or whatever else – so, now it knows.
But I’ve been dabbling. I’ll ask it to find me the cheapest price for something, or the best – based on real reviews, or to turn me into an AI action figure. You know, pretty standard stuff.
My main searches had been for sales and family-friendly recipes. So last week I thought I’d try and combine the two: could AI make me a seven-night, sales-ingredient-only meal plan that the kids would actually eat?
It took me a few goes to get the prompt right. Because if you’ve at all dabbled with whatever iteration, you know it’s all about the prompt. But yes, it can.
First, I specified that each meal must include protein and vegetables. And different types each night (how happy the kids would otherwise have been).
Next, I chose one particular supermarket, and let me be clear this is not an endorsement in any way but you need to say what you want the AI to interrogate.
AI brought the cost of my entire shop down to $106.85.
I tried to get it to search particular web pages for specials: each of Woolworths’ half-price, lower winter price, lower shelf price and online only specials pages. No dice – neither ChatGPT nor Claude could do that with accuracy, nominating typical savings only.
So, I downloaded the current Woolworths sales catalogue and then uploaded it to AI. I wanted to know the precise, I guess, ‘before and after’ prices of all the ingredients. I played with how best to ask for that.
And I needed, naturally, both the nominated meals’ ingredients and instructions. Besides that, a shopping list was pretty key to my convenience. So what meals did it serve up?
- Night 1: Lemon Herb Chicken with Roast Veggies
- Night 2: Beef Stir-Fry with Broccoli and Noodles
- Night 3: Baked Salmon with Green Beans and Potatoes
- Night 4: Pasta with Mince and Zucchini
- Night 5: Vegetable Fried Rice
- Night 6: Chicken and Veggie Tray Bake
- Night 7: Sausage and Bean Casserole
Now, as these delectable dishes emerged on screen, there was a mixed reaction from my progeny. “Salmon is gross.” “Eww, zucchini.”
But, of course, you can add exclusions applicable to your own family, in the prompt. I kept it deliberately varied to cater to as many tastes as possible. And how much – the aim of the exercise – did I save?
Across the week, $38.85, which really downplays the fact that AI – and it should be said I got the best result from the latest free version of ChatGPT – brought the cost of my entire shop down to $106.85.
But I wanted to know about any leftover ingredients, too, so I added specifications to also capture the usage – and discovered my net cost per serve for the week was just $3.06.
Here is the bottom-line summary, as requested:
- Summary of costs: original total price $145.70
- Sales price: $106.85
- Value of leftovers: $21.30
- Net spend (after leftovers): $85.55
- Net cost per serve (28 serves): $3.06
So what was the prompt that made all this food, financial and mental ease happen? This was the exact wording that produced the culinary, cost-cutting goods: “Devise a 7-night dinner plan for four people with no food allergies. Choose family-friendly recipes with a maximum of eight ingredients with varied protein and vegetables in each. Only include recipes that can be made with items that are in the Woolworths sales catalogue, uploaded. List the ingredients and quantities. Make a weekly summary with the following totals: the original, undiscounted price that the ingredients (in the minimum quantity that can be purchased to make the recipe) would have cost; the sales price of those ingredients today; and the monetary value of the ingredients that will be left over after making the recipes, based on the sales price of the minimum quantity that can be purchased today. Calculate the net cost per serve based on the sales price, removing the cost of the leftover ingredients. Include cooking instructions after each recipe. Make a shopping list of the ingredients at the end.”
With that precise prompt (which you can amend for your sales catalogue and food requirements of choice), your menu and money could get a lot easier.
And yep, I had a lovely, categorised shopping list at the end of all the recipe excitement, to make the meal magic happen. ChatGPT even added a question: “Let me know if you’d like suggestions for how to use up those leftovers next week.“
Let the online record show, I now love AI – a lot!
Nicole Pedersen-McKinnon is the author of How to Get Mortgage-Free Like Me, available at www.nicolessmartmoney.com. Follow Nicole on Facebook, Twitter, and Instagram.
- Advice given in this article is general in nature and is not intended to influence readers’ decisions about investing or financial products. They should always seek their own professional advice that takes into account their own personal circumstances before making any financial decisions.
Expert tips on how to save, invest and make the most of your money delivered to your inbox every Sunday. Sign up for our Real Money newsletter.