A Guide to Generative AI

Generative AI has been one of the buzziest topics in tech lately, and for good reason. It’s no longer just a tool for tech giants - small and medium businesses (SMBs) can now tap into it to boost productivity, reach customers in new ways, and save time on tasks that once felt like a grind.

But “generative AI” isn’t one single thing. It comes in different flavours, each with its own strengths and weaknesses. Below, we’ll unpack the main types of generative AI you’re most likely to come across, and weigh up the pros and cons for businesses like yours.

Text Generation (Chatbots, Copywriting, Reports)

These AI models create human-like text. You’ve probably heard of ChatGPT or Claude - they can draft emails, marketing copy, policy documents, or even act as a virtual assistant. In most cases they won’t be able to carry out actions like actually drafting then sending an email automatically - at least not yet.

Positives:

• Speeds up repetitive writing tasks (emails, proposals, product descriptions).
• Can support customer service with 24/7 chat responses.
• Affordable entry point - many tools have free or low-cost tiers.

Negatives:

• Sometimes produces content that sounds right but isn’t accurate.
• Requires oversight - legal or technical details should always be checked.
• May lack the “brand voice” you’ve carefully built.

Image Generation (Logos, Social Media Posts, Product Visuals)

Tools like DALL·E, Midjourney, and Stable Diffusion can generate high-quality images from text prompts. Great for marketing visuals when budgets are tight.

Positives:

• Fast, inexpensive way to get visuals without hiring a designer every time.
• Useful for quick concept mock-ups or seasonal campaigns.
• Can spark fresh creative ideas you might not have considered.

Negatives:

• Not always on-brand - AI designs can look generic or “off” without tweaking.
• Quality varies, and you may still need a human designer to polish.
• Licensing and copyright rules around AI-generated images are still evolving.

Audio & Speech Generation (Voiceovers, Podcasts, Training Content)

Generative AI can now produce realistic speech in different voices and accents. Handy for explainer videos, phone systems, or even turning blogs into audio.

Positives:

• Cuts costs on hiring voice actors for smaller jobs.
• Quick turnaround for audio content.
• Accessible for audiences who prefer listening over reading.

Negatives:

• Can sound a bit robotic if not tuned carefully.
• Ethical questions if voices mimic real people without permission.
• Limited emotional range compared to professional voice talent.

Video Generation (Adverts, Tutorials, Social Clips)

Video AI tools like Runway, Synthesia, or Pika Labs can create explainer videos or short promotional clips from text scripts.

Positives:

• Makes video content more accessible to businesses without big budgets.
• Speeds up production of training or onboarding videos.
• Can help you compete with larger players who push lots of video content.

Negatives:

• Early-stage tech - results can be hit-and-miss.
• Lack of human authenticity can turn viewers off.
• May need editing skills to get professional polish.

Code Generation (Automation, Web Tools, Data Scripts)

For businesses with a bit of tech in their stack, AI can help write code snippets, automate tasks, or build prototypes. Tools like GitHub Copilot or AI assistants in no-code platforms fall into this bucket.

Pros

• Saves time for IT teams or contractors.
• Makes small-scale automation (like spreadsheets, reporting, website tweaks) more achievable.
• Can empower non-technical staff to experiment.

Negatives:

• Risk of introducing bugs if code isn’t reviewed.
• Not a replacement for experienced developers.
• Security concerns if sensitive data is shared with AI tools.

Making the Most of Generative AI in Your Business

For SMBs, the sweet spot is usually using generative AI to save time and stretch budgets, rather than to fully replace skilled humans. Think of it as a junior team member: quick, eager, and sometimes brilliant - but needing a bit of supervision.

It’s extremely important that the output from generative AI is reviewed by a real person who is knowledgeable in the content.

Tips to get started

• Start small. Try AI for internal tasks first (drafting reports, creating mock-ups).
• Keep human oversight. Always review AI work before sharing with clients or the public.
• Match the tool to the job. Don’t use image generators for legal contracts or chatbots for complex customer service.
• Watch the costs. Many platforms charge per use - keep an eye so you don’t rack up surprise bills.

Final Word

Generative AI isn’t just for Silicon Valley - it’s becoming a practical tool for businesses from Tauranga to Te Awamutu. Whether it’s polishing up your marketing, cutting down admin time, or helping you experiment with new ideas, there’s real potential.

But use any AI generated content with care. AI “slop” is a term describing the huge amount of content that might look OK but is actually incorrect or of low value. An ironic view of technical AI content is that the more technically literate people will get more value as they know what’s right and wrong.

Just remember: it’s not about replacing people - it’s about freeing them up to focus on the work that really matters. AI’s not coming for your job, it’s coming to your job.