I gave a talk at WordCamp Europe 2020 (online, because pandemic) about something that felt simultaneously overhyped and underutilized: Artificial Intelligence in WordPress.
This was June 2020. Pre-ChatGPT. Pre-LLM explosion. Pre-"AI will replace all programmers" discourse. The AI landscape was very different – lots of narrow, specialized tools rather than one general-purpose assistant.
The challenge: It’s hard to tell hype from genuine utility. AI vendors promise magical solutions. Skeptics dismiss everything as snake oil. The truth is messy and specific.
So I focused on practical, achievable ways to integrate AI into WordPress tech stacks. Not "AI will revolutionize everything" but "here are nine specific things you can do today."
So yeh. Let me walk you through what I covered.
The Context: AI Is a Bundle of Technologies
AI isn’t one thing. It’s a bundle of over 25 different technologies, ranging from everyday to cutting-edge.
Everyday AI (you use it constantly without thinking):
- Spam filters (machine learning classification)
- Autocorrect (natural language processing)
- Recommendation engines (collaborative filtering)
- Image recognition (computer vision)
- Voice assistants (speech recognition + NLP)
Cutting-edge AI (research labs, specialized applications):
- Generative adversarial networks (GANs)
- Reinforcement learning
- Transfer learning
- Transformer models (GPT-3 had just launched)
Most WordPress users don’t need cutting-edge. They need practical applications of mature AI technologies. That was my focus.
Three Categories, Nine Examples
I organized the nine ways into three categories:
Category 1: AI Enhancement
Using AI to enhance user experiences and content quality.
1. Automated Image Alt Text
Tools like Microsoft Azure’s Computer Vision API or Google Cloud Vision can analyze images and generate descriptive alt text automatically.
Why it matters: Accessibility compliance. SEO benefits. Saving hours of manual work.
WordPress implementation: Plugins that integrate these APIs, auto-generating alt text on image upload. Not perfect, but 80% accurate is better than missing entirely.
2. Content Recommendations
Amazon-style "you might also like" for blog content. Machine learning analyzes user behavior, suggests relevant posts/products.
Why it matters: Increased engagement. Lower bounce rates. Better user experience.
WordPress implementation: Plugins using collaborative filtering or content-based recommendation algorithms. Some self-hosted, some cloud-based.
3. Automated Content Tagging
Natural language processing analyzes post content, suggests tags/categories automatically.
Why it matters: Better content organization. Improved internal linking. Time savings for authors.
WordPress implementation: NLP APIs (AWS Comprehend, Google Natural Language) integrated via plugins. Suggests tags based on content analysis.
Category 2: Improved Productivity
Using AI to reduce repetitive work and automate tasks.
4. Chatbots for Customer Support
AI-powered chatbots handle common questions, escalate complex issues to humans.
Why it matters: 24/7 support availability. Reduced support ticket volume. Faster resolution for common queries.
WordPress implementation: Many WordPress-compatible chatbot platforms (Tidio, Drift, ManyChat). Some with sophisticated NLP, some with simple rules.
5. Automated Content Generation
AI tools generating product descriptions, social media posts, email subject lines.
Why it matters: Content at scale. Consistent brand voice. First drafts that humans refine.
WordPress implementation: In 2020, this meant tools like Copy.ai (just launching), Jasper (not yet existed), or GPT-3 API integration (very new). Primitive compared to 2023+ capabilities.
6. Automatic Translation
Neural machine translation (Google Translate, DeepL) for multilingual WordPress sites.
Why it matters: Global reach without manual translation costs. Accessibility for international audiences.
WordPress implementation: Plugins integrating translation APIs. Not perfect, but good enough for understanding. Human post-editing for quality.
Category 3: Data, Analytics, and Intelligence
Using AI to understand your content, users, and performance.
7. Predictive Analytics
Machine learning analyzing user behavior patterns, predicting churn, conversion likelihood, content performance.
Why it matters: Data-driven decisions. Early intervention for at-risk users. Content strategy optimization.
WordPress implementation: Analytics tools with ML capabilities (Google Analytics with predictive metrics, specialized WooCommerce analytics plugins).
8. Sentiment Analysis
Natural language processing analyzing comments, reviews, social mentions to gauge audience sentiment.
Why it matters: Understand how content is received. Identify PR issues early. Measure brand perception.
WordPress implementation: Comment analysis plugins, integration with social listening tools, API connections to sentiment analysis services.
9. SEO Optimization
AI tools analyzing top-ranking content, suggesting keyword strategies, identifying optimization opportunities.
Why it matters: Better search rankings. Data-driven content strategy. Competitive intelligence.
WordPress implementation: SEO plugins with ML features (Yoast, Rank Math), dedicated AI SEO tools (Clearscope, Surfer SEO integrating with WordPress).
The Practical Reality
These weren’t theoretical. In 2020, all nine had available tools/plugins/services. Some expensive, some affordable, some free.
The challenge wasn’t "can you do this?" but "should you, and how much effort is it worth?"
Cost-benefit analysis example:
Automated alt text: High benefit (accessibility, SEO), low cost (API calls pennies per image), easy implementation (plugins exist). Do it.
Predictive analytics: High benefit (if you have scale), high cost (data infrastructure, analysis), moderate complexity. Maybe, if you’re WooCommerce with thousands of customers.
AI content generation: Moderate benefit (saves time), moderate cost (API usage or subscriptions), high risk (quality issues, brand voice). Experiment carefully.
What’s Changed Since 2020
Looking back from 2025, this talk feels simultaneously prescient and dated.
What was right:
- AI would become standard in WordPress workflows
- Practical, narrow applications beat general hype
- Cost-benefit analysis matters more than technological capability
- Integration challenges (APIs, plugins, UX) remain barriers
What changed dramatically:
- LLMs (ChatGPT, Claude) collapsed multiple narrow AI tools into single general-purpose assistants
- Content generation went from "interesting experiment" to "mainstream tool" overnight
- Cost of AI dropped orders of magnitude (GPT-3 in 2020 vs GPT-4 in 2024)
- Quality of AI outputs improved far faster than predicted
In 2020, I suggested nine specific, narrow applications. In 2025, you could arguably handle most of them with ChatGPT or Claude plus WordPress integration.
That’s both exciting (lower barriers) and concerning (overreliance on black-box systems).
The Core Message Remains
Despite technological change, the core message still holds:
AI is a tool, not magic. Understand what it’s good at. Understand where it fails. Apply it where cost-benefit makes sense.
Start with problems, not technology. Don’t implement AI because it’s trendy. Implement it because you have a specific problem it solves better than alternatives.
Hype vs. utility distinction matters. Vendors will promise the moon. Skeptics will dismiss everything. The truth is always specific and contextual.
Augmentation beats replacement. AI enhances human capabilities; it doesn’t (yet) replace human judgment, creativity, and strategic thinking.
Those principles applied in 2020. They apply in 2025. They’ll probably apply for a while.
Why This Talk at This Time
WordCamp Europe 2020 was fully online (pandemic). That changed the format – pre-recorded talks, smaller audience engagement, different energy.
But it also meant global reach. The talk reached WordPress users worldwide who might not have traveled to Porto for an in-person conference.
AI in WordPress was still niche in 2020. Most site owners weren’t thinking about it. Developers were experimenting. Agencies were pitching it to clients without always understanding it.
The talk aimed to demystify – here’s what AI actually means, here’s what’s practically possible, here’s how to evaluate whether it makes sense for your use case.
Not revolutionary. Just useful, practical information at a time when AI was transitioning from research labs to mainstream tools.
Event: WordCamp Europe 2020 Online
Talk: Nine Ways to Make WordPress Better With AI
Date: June 18, 2020
Duration: 35:21
Watch: WordPress.tv
License: Creative Commons Attribution-ShareAlike 4.0 International
Find me: Contact form, @divydovy, hi@divydovy.com
Note: This talk was given in June 2020, before the explosion of LLMs and generative AI that started with ChatGPT in late 2022. The AI landscape has changed dramatically since then, but the principles about evaluating utility vs. hype remain relevant. The specific tools and APIs mentioned are now often replaced by more general-purpose LLM integrations.
Sources:
- WordPress.tv – Nine ways to make WordPress better with AI
- SlideShare – Using off-the-shelf AI tools to augment WordPress
- Angry Creative – WordPress & AI
