SkillHack › Career guides › AI Research Scientist
How to become an AI Research Scientist
Novel architectures, training methods, scaling laws, and publication. Push the frontier of what models can do.
- 5phases in the roadmap
- 15topics to work through
- 75graded practice questions
- Freeno payment, ever
The AI Research Scientist roadmap
Each topic below carries a study note, real reading, and practice questions in the app. A topic unlocks when you pass the one before it, so the order is the path, so you are never guessing what to learn next.
Phase 1AI research science foundations
The scientific method in ML
Machine learning research is still science: an observation is a hypothesis, not a conclusion.
Navigating the literature
New ideas rarely have no ancestors.
Choosing research problems
What you choose to work on bounds the impact of everything after.
Phase 2AI research methods
Designing experiments
An experiment can only attribute an effect to a cause when everything else is held equal.
Hypotheses and baselines
A claim is only as strong as the baseline it beats.
Ablation studies
When a method bundles several new pieces, the headline number cannot tell you which piece matters.
Phase 3AI research science modeling
Architectures
Architecture choice is a hypothesis about the structure of your data.
Training dynamics
When training destabilizes, guessing is expensive.
Scaling laws
Scaling laws let you predict a large run's performance and allocate compute between model size and data before committing.
Phase 4AI research science evaluation
Rigorous evaluation
A benchmark number is only as trustworthy as the gap between its data and your training set.
Statistical significance
A small gap on one seed can be pure randomness.
Reproducibility
Work nobody else can rerun is a claim rather than a contribution.
Phase 5AI research science communication
Writing papers
A paper's job is to make precisely the claims its evidence supports.
Peer review
Good reviewing judges work on evidence and rigor, not on the prestige of the author list.
Research impact
Impact is not the same as attention.
Reading for this path
The primary sources behind the topics above, all free to read.
- Troubling Trends in ML Scholarship (Lipton & Steinhardt)
- A Recipe for Training Neural Networks (Karpathy)
- How to Read a Paper (Keshav)
- Semantic Scholar (citation graph)
- You and Your Research (Hamming)
- How to do great research (Anthropic careers)
- Deep Reinforcement Learning that Matters
- Show Your Work: Improved Reporting of Experimental Results
- On the Importance of Baselines (Dacrema et al.)
- A Metric Learning Reality Check
- Ablation Programming for ML (analysis)
- The Illustrated Transformer (component intuition)
- Attention Is All You Need
- The Inductive Bias of ML (Battaglia et al.)
Coming from another job?
Most people on this path arrived from somewhere else: engineering, analysis, testing, product, support, compliance, design. Onboarding asks what you do today and builds a short starter run out of the gaps, so you begin from what you already know rather than from chapter one. See how it works.
Other AI career paths
AI Safety Engineer
Risk analysis, red teaming, privacy, policy, evaluation design, and monitoring. Probe, evaluate, and govern AI systems.
AI Security Engineer
Prompt injection defense, model supply chain, data leakage, and access control. Secure AI systems end to end.
AI Solutions Architect
System design, model selection, build-vs-buy, and scaling and cost. Architect enterprise AI solutions that hold up in production.
AI Strategist
Opportunity sizing, build-vs-buy, and adoption strategy. Decide where AI creates real business value.
AI Systems Engineer
GPUs, distributed training, kernels, and memory and throughput optimization. Make large models train and serve fast.
AI UX Designer
Human-AI interaction patterns, trust, uncertainty, and feedback loops. Design interfaces where people and models work well together.
Start the AI Research Scientist path for free
The full study notes, the reading, and the practice questions behind every topic above are in the app. Answer one tonight and you have started.
No payment, no credit card, no CV. Sign in with Google.