Blackcurrant

Podcast 2 — In conversation

Speaker A: I made a blackcurrant and apple crumble last night and it came out brilliantly. The blackcurrants give it this wonderful sharp depth that you just do not get from any other berry.

Speaker B: Blackcurrant is one of those flavours that is distinctly British in a way that is hard to explain to someone who did not grow up with Ribena. It is almost nostalgic before you have even tasted it.

Speaker A: Ribena is exactly what I mean. That deep, slightly medicinal sweetness — I loved it as a child.

Speaker B: And it is not just nostalgia — blackcurrants really do have exceptional nutritional properties. The vitamin C content is remarkable for a soft fruit.

Speaker A: Now, is blackcurrant always one word? I sometimes see it written as two — black currant.

Speaker B: In British English, the standard form is one word: blackcurrant. American English tends toward two words — black currant — partly because the fruit is less common there and the compound has not solidified in the same way. If you are writing for a British audience, always one word.

Speaker A: And what about close relatives? Redcurrant and whitecurrant — are those all the same fruit in different colours?

Speaker B: Related but distinct varieties of Ribes. Redcurrants are smaller and much sharper, used mainly in jellies and sauces rather than eaten fresh. Whitecurrants are milder and sweeter, less common commercially. The blackcurrant is by far the most intensely flavoured of the three.

Speaker A: What is the difference between a blackcurrant and a blueberry, which some people might confuse?

Speaker B: Very different in flavour. Blueberries are mild, sweet, slightly floral. Blackcurrants are tart, astringent, almost peppery in intensity. They look superficially similar — small, dark, round — but the taste is completely distinct. You would not mistake one for the other after tasting both.

Speaker A: And cassis — the French liqueur. That is blackcurrant as well, isn't it.

Speaker B: Crème de cassis, yes — blackcurrant liqueur from Burgundy. Used in a kir or a kir royale. The word cassis is simply the French for blackcurrant. So when you order a kir royale, you are essentially drinking champagne and blackcurrant.

Speaker A: Which is an excellent way to spend an afternoon. Right — blackcurrant. One word in British English, intensely flavoured, very British, with a French alias of cassis.

Speaker B: Perfectly summarised. A small berry with an outsized personality.

Podcast 3 — Prompt engineering

Speaker B: Right — blackcurrant in development prompts is primarily a food and beverage domain word, but it is a very rich one: flavour profiles, nutritional data, recipes, product catalogues, and colour palettes all come into play. Let us work through the practical prompt applications.

Speaker B: Starting with a product catalogue for a food business.

Build a product catalogue page for a British artisan jam and preserve company. Display products in a responsive grid. Each product card shows the product name, flavour — such as blackcurrant and apple, or blackcurrant and elderflower — net weight, price, and a brief tasting note. Include a filter sidebar beside the grid to filter by fruit type, with blackcurrant as a primary category.

Speaker B: Blackcurrant as a primary filter category tells the AI this is a significant product line — not a one-off. It generates a filter with weighted categories rather than a flat alphabetical list. The flavour combinations — blackcurrant and apple, blackcurrant and elderflower — are real product names that produce realistic demo content rather than placeholder text.

Speaker A: Filter sidebar beside the grid is doing layout work again. Left sidebar, right product grid — without beside, the AI might put the filter above the products in a mobile-first stacked layout.

Speaker B: One word, correct layout. Now a recipe platform.

Create a recipe detail page for a food and drink platform. The recipe is blackcurrant cheesecake. Show: ingredient list with quantities, step-by-step method numbered from one, preparation time, servings, and a nutritional information panel. Highlight the word blackcurrant wherever it appears in the method steps. Include a reviews section below the recipe with star ratings beside each review.

Speaker B: Highlight the word blackcurrant wherever it appears is a specific interactivity request that produces a text-scanning function — the AI wraps matching terms in a styled span rather than building a static page. Star ratings beside each review gives the rating its correct placement: inline with the reviewer name, not below it.

Speaker A: That highlight function is actually a useful vocabulary learning mechanic — students can see the target word in real cooking context. I would not have thought to request that.

Speaker B: Good observation. Combining the word with its natural domain always enriches the output. Now a nutritional database.

Design a PostgreSQL schema for a nutritional ingredients database used by a food technology company. Tables: ingredients with id, common_name, latin_name, and category; nutrients with id, name, unit; ingredient_nutrients with ingredient_id, nutrient_id, amount_per_100g, and data_source. Populate seed data for blackcurrant including vitamin C at 181 milligrams per hundred grams, anthocyanins, and potassium. Index on ingredient_id.

Speaker B: Vitamin C at 181 milligrams per hundred grams is a real nutritional figure for blackcurrant. Providing accurate domain data in the prompt produces accurate seed data in the output — not invented placeholder values. The anthocyanins entry also signals that the schema should handle named phytonutrients, not just macro and micronutrients.

Speaker A: Data_source field is a clever addition — nutritional databases always need to cite where figures came from because values vary between different laboratory analyses.

Speaker B: Provenance is critical in scientific data. Always include it. Now a drinks menu system.

Build a drinks menu management system for a bar and events company. Cocktails have a name, base spirit, mixer list, garnish, glass type, and price. Create three blackcurrant cocktail entries: a kir royale with champagne and blackcurrant liqueur, a blackcurrant mojito with rum and blackcurrant cordial, and a blackcurrant sour with gin and fresh blackcurrant puree. Display the menu with cocktail descriptions beside each drink image placeholder.

Speaker B: Three named cocktail entries with real ingredients remove all ambiguity from what the system needs to store. The AI generates exactly the right field structure — base spirit separate from mixer separate from garnish — because the real-world examples demonstrate the data shape better than a schema description alone.

Speaker A: Descriptions beside each drink image — I keep noticing how much layout work that word beside does. It is the most efficient layout word in prompt writing.

Speaker B: It eliminates a whole paragraph of layout description. Final two prompts — colour and e-commerce. First, colour.

Generate a CSS colour palette for a premium British food brand whose signature product is blackcurrant preserve. The palette should include: a deep blackcurrant purple as the primary brand colour, a warm cream as the background, a berry red as an accent, and a dark charcoal for body text. Provide CSS custom properties for each colour with descriptive variable names like --color-blackcurrant and --color-cream.

Speaker B: The blackcurrant colour specification drives the entire palette derivation. The AI starts from the fruit's actual deep purple and generates harmonious complementary colours rather than picking arbitrarily. Descriptive variable names like --color-blackcurrant make the CSS self-documenting — any designer reading it immediately knows the brand intent.

Speaker A: And the cream background rather than white signals premium artisan positioning. That one word — cream — shapes the brand register of the whole palette.

Speaker B: Vocabulary is branding. Now the e-commerce prompt.

Create a product detail page for an online farm shop selling fresh blackcurrants. Include: product title, price per kilogram, stock status, weight selector from 250g to 2kg, add to basket button, origin information — UK-grown, June to August season — storage instructions, and a suggested uses section listing jam, cordial, crumble, and smoothies. Display customer reviews beside a summary star rating at the bottom.

Speaker B: UK-grown, June to August season is real horticultural data that produces accurate, trustworthy product copy. Storage instructions and suggested uses are genuine e-commerce conversion elements — they answer the customer questions that prevent a purchase from being abandoned.

Speaker A: The weight selector from 250g to 2kg gives the AI both the UI component — a stepper or dropdown — and its range. Without that range, the AI invents arbitrary options.

Speaker B: Always bound your ranges in prompts. A blackcurrant is a small thing. A well-specified prompt is not.