AI, But Simple

The AI intelligence layer.

Real interview questions, job trends, top papers and code. PRO watches hundreds of sources and ranks what matters.

One subscription, three tracks.

The visual deep dive

Weekly ideas, visually explained with code. Read deeper insights and obtain exclusive issues.

Inside the intelligence layer.

Interview tracker

The questions companies actually ask.

Sourced from over 100 companies with answers and working code, grouped by topic.

PRACTICE SET FULL BANK · 627 → ML FUNDAMENTALS · 35 How do you combat the curse of dimensionality? ML FUNDAMENTALS · SOURCED · ANDREWEKHALEL/MLQUESTIONS What's the trade-off between bias and variance? ML FUNDAMENTALS · SOURCED · ANDREWEKHALEL/MLQUESTIONS Simple models miss structure, complex models memorize noise. The sweet spot minimizes total error on data the model has never seen. error = bias**2 + variance + noise REVIEW SOLID What is gradient descent? ML FUNDAMENTALS · SOURCED · ANDREWEKHALEL/MLQUESTIONS Explain over- and under-fitting and how to combat them.
Job trends

Demand, quantified.

4,400+ live postings, ranked by role and company demand, re-scored every day.

DEMAND BY ROLE Sales & GTM 611+13 Solutions & forward-depl. 409-4 Infrastructure & platform 301+13 Agent roles 189+8 Product & design 151+3 Research engineer 81+3 Inference & serving 71+1 Safety & alignment 66-1 Training & pre-training 48+1 POSTINGS BY COMPANY Databricks500 OpenAI500 Anthropic433 Harvey352 Palantir281 xAI230
Pulse

The week's signal, grouped.

Releases, agents, benchmarks, funding: the news that matters, clustered by topic and ranked by where the signal came from.

All topics · 100 Models & releases · 19 Agents · 6 Coding & SWE · 5 Reasoning MODELS & RELEASES A 7B open model closes the gap on coding evals NEW arxiv.org · 2h ago Ask HN: what breaks when you fine-tune on synthetic data? 87 pts · news.ycombinator.com · 1d ago Agent frameworks converge on a shared tool protocol github.com · 5h ago Serving costs fall as quantized inference goes default TOP SOURCES arxiv.org 28 github.com 21 huggingface.co 14 news.ycombinator.com 11 openai.com 7

One layer. Eight systems.

No. 1

Interview question tracker

Hundreds of sourced questions with answers and working code.

No. 2

Paper tracker

The research that moved each week, ranked and decoded.

No. 3

GitHub & model tracker

Trending repos and weights, attached to their papers.

No. 4

Job trends

4,400+ live postings, ranked by role and company demand.

No. 5

Code vault

Every code notebook from the newsletter, organized by topic.

thecatsatonthemat thecatsaton HEAD 2 · SYNTAX CHECK YOUR WORK Drag the head. The pattern moves with you.
No. 6

Interactive articles

Beautiful interactive figures for enhanced understanding.

No. 7

Roadmaps & career paths

Fundamentals to offer, built from the live question bank.

No. 8

Pulse

The week's AI news, clustered by topic and ranked.

Simple, like the newsletter.

Free
$0
Every other week, forever.
  • The visual deep dive, biweekly
  • Limited archive
  • Read on the web or by email
Subscribe free
PRO2 mo free
$10 / month, billed annually
The AI intelligence layer + more deep dives.
  • Everything in Free
  • Interview tracker, sourced from 100+ companies
  • Paper tracker & GitHub/model tracker
  • Code tutorial with every issue (Code vault)
  • All interactive articles
  • Roadmaps & career paths
  • Ad-free issues
Upgrade to PRO

Questions, answered simply.

What stays free? +

The visual deep dive, forever: biweekly on Free. PRO steps it up to weekly, opens the full archive, and adds the intelligence layer, the interactive articles, and the roadmaps on top.

How is this different from following arXiv or another newsletter? +

A feed shows you everything and ranks nothing. PRO watches the whole field, shows what actually moved, and connects the papers, repos, postings, and interview questions to each other, then narrows it to what matters for the work you do.

How often do the trackers update? +

Continuously. The pipeline pulls from hundreds of sources every day and re-ranks on each run, so what you see reflects the current state of AI.

How is the data selected? +

We monitor hundreds of sources continuously, then a proprietary ranking algorithm using 50+ factors decides what surfaces. Items are attributed: every interview question carries its company, role, and when it was asked, plus an answer and working code where it applies. Your preferences decide what leads.

Get started with PRO.

Start now

or read a free issue first