Working with Modern AI
Understand how contemporary artificial intelligence models actually process, reason, and generate information. Learn to steer models via structured outputs, ground them with retrieval systems, and navigate the architectural trade-offs powering modern AI applications.
Like Duolingo, but for Working with Modern AI. Tomo turns the whole topic into a game you play five minutes a day, until it actually sticks.
25 levels across 3 sections, about 50 minutes end to end, roughly 10 days at five minutes a day. It moves through Controlling Model Behavior; Extending AI with External Data; and Systems, Reasoning, and Inner Mechanics. It assumes you already know the basics.
Free forever · No credit card · iPhone & Android
Key ideas in Working with Modern AI
- System prompts hold elevated steering priority across all dialogue turns.
- User turns supply task inputs that can be overridden by system framing.
- Assistant turns prime the AI by mocking desired prior responses in history.
- Few-shot examples anchor structural formatting better than rules alone.
- Autoregressive patterns mirror concrete input-output syntax cues.
- Abstract descriptions frequently drift on subtle edge-case formatting.
- Raw logits convert to probabilities across the entire active vocabulary.
- Top-k drops everything outside the highest fixed number of candidates.
- Top-p trims the remaining tail once cumulative probability is reached.
- Each generated intermediate token serves as attention context for next tokens.
- Forcing direct answers starves complex multi-step reasoning of compute.
- Models cannot revise an emitted token without dedicated intermediate steps.
- Specifying positive substitute behaviors prevents priming the forbidden word.
- Applying API logit bias directly eliminates forbidden token probabilities.
- Few-shot demonstrations show correct phrasing without naming the bad token.
- The engine masks invalid vocabulary tokens at each generation step using the schema grammar
You've tried the other tabs
Thirty open tabs. Four facts you actually kept.
You watched. You nodded. By Sunday it was gone.
One answer, then back to scrolling.
Eight weeks. You meant to finish. You didn't.
Tomo gives Working with Modern AI the Duolingo treatment: levels, streaks, and quick quizzes that test what you just learned. That game loop is what the tabs above never had, so it's the one you actually finish.
Here's what playing it feels like
A real question from this course. Take your best guess.
What happens when a user's input directly clashes with rules set in the system prompt?
Get it right to open this lesson and 24 more in the app.
Where Working with Modern AI takes you
- 1
Controlling Model Behavior
- Prompt Architecture and Steering Mechanics
- Function Calling and Structured Outputs
- 2
Extending AI with External Data
- Vector Embeddings and Semantic Search
- Retrieval-Augmented Generation Architecture
- Adapting Models: Context vs Fine-Tuning
- 3
Systems, Reasoning, and Inner Mechanics
- Agent Workflows and Multi-Step Reasoning
- Inside the Transformer Engine
- Alignment, Guardrails, and Vulnerabilities
3 sections · 8 units · 25 levels. Built to play, not to enroll.
Tomo is in ChatGPT
Tomo is also a ChatGPT plugin: the same playable courses, built and played inside your chat.
Add Tomo to ChatGPT- What to ask
- “Make me a Tomo course on black holes.” “Turn this chat into a Tomo course.”
- What you get
- Quick questions with instant feedback, hearts and stars, ready to play in about a minute.
- Where it goes
- No sign-up to start. Connect your Tomo account and your streak carries on in the app.
You pick the voice
Working with Modern AI is taught in the Bestie style: your friend who just gets it. Want a different feel? In the app you can spin up the same topic in any of Tomo's teaching styles. Same facts, totally different vibe.
More Technology on Tomo
The Modern AI Stack
Move beyond basic prompt engineering to understand how cutting-edge AI systems actually operate. Explore the architectural shifts powering retrieval, tool-calling agents, inference scaling, and multi-step autonomy.
Practical AI and Prompt Workflows
Move beyond conversational chat to use language models as analytical engines and structured workflow partners. Learn prompt architecture, token mechanics, schema extraction, and tool integration.
AI-Native Software Engineering
Transition from using AI as a chatbot to integrating it as a core architectural component. This course covers advanced agentic workflows, automated evaluation loops, and the shift toward non-deterministic system design.
Dynamic Excel Workflows
Upgrade from fragile cell formulas to resilient, modern spreadsheet design. Learn how to leverage dynamic arrays, flexible lookups, and structured tables to build spreadsheets that never break when data updates.
Advanced ChatGPT Workflows
Shift from treating ChatGPT as an unpredictable oracle to operating it as a steerable, deterministic cognitive engine. Master structural prompt architecture, adversarial verification protocols, and reproducible execution pipelines.
Claude for Power Users
Unlock the full engineering and analytical capabilities of Anthropic's Claude. Learn to control model reasoning with structured prompting, build reliable agentic systems using Model Context Protocol, and orchestrate complex autonomous workflows at scale.
Start Working with Modern AI today.
Download Tomo, search Working with Modern AI, and play your first lesson in under a minute.