AI is coming for your job unless you arm yourself with the right skills.
Here is a clear guide to learning AI, from complete beginner to advanced level.
1. Generative AI Literacy
If you don’t yet understand what generative AI is, and how it works, start here. Gen AI is everywhere now (emails, images, videos, chatbots). Google’s AI Essentials course is a good place to start. Checkout Grow with Google.
2. Prompt Engineering
Prompting is the language of AI models. Master prompting and you’re ahead of 98% of people.
Remember two frameworks you can use when constructing your prompts:
TCR EI → Task, Context, References, Evaluate, Iterate
RSTI → Role, Steps, Tone, Input
Good prompting unlocks every other AI tool. It’s the swiss army knife of mastering AI.
3. Specialized AI Tools
You don’t need to know a tonne of tools. Pick one general chatbot (ChatGPT, Claude, Gemini, Perplexity) and go deep.
Must have capabilities:
• Answer questions & analyze images/data
• Research & personalized news
• Supercharge learning
• Build slides/dashboards/simple apps
4. AI Agents
Companies are building tailored agents (retention, reporting, etc.) that integrate into existing workflows.
Learn to building custom AI agents with no-code or code frameworks (n8n, make, etc). This is one way to stand out in your current job because of the huge demand internally + freelance.
5. Open-source AI
Closed-source models like OpenAI's GPTs, and Google's Gemini dominated… until Chinese open-source models (DeepSeek, Qwen) matched performance.
Open-source models are cheaper, provide full control, customizable, no vendor lock-in, auditable (key for finance, healthcare and law enforcement).
This year will see massive shift to open-source. Western companies will open-source more too. Lower cost + more innovation for everyone. Learn how to deploy them.
6. AI-Assisted Coding (“Vibe Coding”)
Anyone can build products now.
Describe what you want → AI builds it. Tweak → ship.
Consider Claude code, Copilot and Cursor.
You may not be able to learn how to build world class scalable apps, but learn to build micro apps and tools that improve your productivity.
Start small, practice daily, and compound your skills. The future belongs to those who adapt now. Arm yourself with AI. Your job (and career) depends on it.
@OloyeSomorin Jeep wrangler user here🙋♂️
The brainbox is one of the worst issues a Jeep can experience, sourcing it is no the problem, programming it is. My experience.
@Ollyskay@Wizarab10 You’ll understand how worthless wealth is without having your loved ones enjoy it. What’s the point of having resources when you can’t help the ones you care about?
Help with every opportunity. Enjoy the journey, you might not get to the destination.
Introducing Perplexity Computer.
Computer unifies every current AI capability into one system.
It can research, design, code, deploy, and manage any project end-to-end.
We’ve identified industrial-scale distillation attacks on our models by DeepSeek, Moonshot AI, and MiniMax.
These labs created over 24,000 fraudulent accounts and generated over 16 million exchanges with Claude, extracting its capabilities to train and improve their own models.
The companies that succeed in the future are going to make very heavy use of AI. People will manage teams of agents to do very complex things.
Today we are launching Frontier, a new platform to enable these companies.
I tried out OpenClaw (Clawdbot, moltbot) and compared it with Claude Cowork.
My personal take: one is your all round 24/7 errand boy, while the other is a polished desk worker. If you’ve been wondering what the buzz is about, here is a summary👇
Clawdbot, moltbot, and OpenClaw are basically the same project with different names over a very short, chaotic few weeks in early 2026.
It's an open-source AI agent/personal assistant. The core idea is simple but powerful:
- runs locally on your computer
- connects to LLMs APIs
- you talk to it through apps you already use: Telegram, WhatsApp, Discord, etc.
- unlike normal chatbots, it can actually do things on your computer like open browsers, click around, read/write files, run apps and terminal commands, automate workflows.
While Claude cowork is an official Anthropic product. It’s not open-source.
It's the "non-coding" evolution of Claude code, same agentic engine, but wrapped in a friendly GUI inside the Claude desktop app.
Both products are powerful use cases of using LLMs to build highly impactful AI tools. We’re still early into 2026, can’t wait to see how they both evolve.