What’s New: LLM-Eliza — A Plugin for Modern LLMs
Recently, a project called LLM-Eliza popped up on Hacker News. It’s a plugin that revives the iconic 1960s chatbot ELIZA by making it accessible via today’s large language models (LLMs). Hosted on Codeberg, this plugin offers a fresh take on the classic AI “therapist” — one of the very first chatbots in tech history.
In short: with LLM-Eliza, anyone using LLMs (like ChatGPT or similar tools with plugin support) can experience a conversation with a virtual “doctor” ELIZA — and see what it felt like to chat with AI half a century ago.
A Quick History: What Is ELIZA?
ELIZA was the first chatbot program, created in 1966 by Joseph Weizenbaum at MIT. It simulated a Rogerian psychotherapist, responding to your statements with questions and rephrased answers.
A classic example:
— I feel bad.
— Why do you feel bad?
ELIZA worked using simple templates and keywords, but it made a huge splash — many people genuinely felt like they were talking to a real, intelligent entity.
Why Modern AI Creators and Marketers Should Care
1. Inspiration and Perspective
LLM-Eliza lets you see where conversational AI truly began. For creators, it’s a chance to look back and appreciate how far human-machine interaction has evolved.
Understanding the roots helps you spot which communication patterns stood the test of time, and which didn’t. It’s a goldmine for brainstorming new chatbot ideas and digital character concepts.
2. UX & Dialogue Design Experimentation
ELIZA is a masterclass in minimalist dialogue. Study why simple reflections sometimes work just as well — or better — than today’s “smart” LLM responses. This is especially valuable if you’re building:
- Customer support chatbots
- AI marketing assistants
- Personal productivity bots
Often, users care less about “perfect” answers, and more about feeling heard — something ELIZA did remarkably well.
3. Contrast & Testing: Past vs. Present
LLM-Eliza gives you a unique way to compare conversations with classic ELIZA and a modern LLM. What’s different? How do users react? Great for A/B testing:
- Run focus groups: which style does your audience prefer — “old-school” or modern?
- Use ELIZA as a retro-style bot to boost engagement in marketing campaigns or storytelling.
4. Nostalgia and Storytelling
Nostalgia is powerful. Marketers can leverage ELIZA’s style and imagery for viral campaigns, creative newsletters, or quirky landing pages.
If your audience loves tech history and unique stories, introducing ELIZA is a proven engagement hook.
5. Ethics, Safety, and Learning
ELIZA never filtered out harmful content, issued safety warnings, or managed sensitive topics — it simply reflected input. Comparing ELIZA with today’s LLMs highlights just how far we’ve come on AI ethics and safety. It’s a vivid teaching tool for:
- AI ethics training sessions
- Webinars for newbie chatbot developers
- Demonstrating “what not to do” versus modern best practices
How to Test LLM-Eliza
Installation is simple (see Codeberg for details), and you can run ELIZA right from your LLM interface. Anyone familiar with LLM plugins — or running them locally — can get started in minutes.
Practical Uses for Business and Creativity
- Education: Illustrate how AI has evolved and spark discussions about changing user expectations.
- Content Marketing: Use ELIZA dialogues as unique content for social media, blogs, or newsletters.
- Internal Training: Compare ELIZA’s “behavior” to modern assistants in common scenarios.
- Gamification: Add ELIZA as an “Easter egg” in your product to surprise and delight users.
If you want to go deeper with LLM and bot-building tools, check out these courses for hands-on guides and integration examples.
What to do right now
- Install and try LLM-Eliza — Chat with it and compare your experience to ChatGPT or other LLMs.
- Analyze ELIZA’s patterns — What conversational tactics does it use? Which still work today?
- Test “retro-style” interactions with your audience — Try it in your newsletter, social media, or landing pages.
- Use ELIZA for training — Show your team AI’s progress from the 1960s to now.
- Explore creative integrations — Think about ways to add a retro chatbot twist to your own projects.
Based on material from Hacker News



