Interview, Explained
In-depth breakdowns of tech interviews — section-by-section analysis with diagrams, quotes, and insights.
~7 minAndrew Feldman on why fast tokens cost more than slow ones
Cerebras says fast inference commands premium token pricing, while AWS, Oracle, and Microsoft give its compute service near-global enterprise reach.
~46 minAmi Luttwak on why the best defense is now to attack yourself
Wiz co-founder Ami Luttwak says 'Black October' proved AI can now find a flaw and run a full attack in about 20 minutes, open-source models are closing the gap with the frontier, and the only survivable defense is to attack your own systems continuously with the latest models.
~18 minChris Lattner on why one AI stack can span every chip
Chris Lattner explains why AI's hardware is fragmenting into specialized chips, how one software stack can span them by unifying what is common and specializing only the hot path, why selling Modular to Qualcomm was what finally let the platform go open source, and where he still keeps a human in the loop as agents take over the code.
~39 minMax Jaderberg on why drug discovery must get out of the real world
Max Jaderberg argues drug discovery stalled because it is still an empirical, real-world trial-and-error process — and the fix is building simulation models accurate enough that candidates are designed and iterated on a computer, leaving the lab only to validate.
~8 minNaveen Rao on why NVIDIA needs an operator, not a CFO
Naveen Rao argues NVIDIA's next CEO must be a deep-technology operator who can make long-horizon bets and read markets - not a finance-first leader - because he sees the CUDA-and-ecosystem advantage that built NVIDIA as what a numbers-only successor could put at risk.
~20 minLisa Su on Why AI Demand Still Outruns Supply
In her CEO-of-the-Year interview, AMD's Lisa Su argues the AI moment is a leadership test: lead by learning when no one has the answers, keep plans dynamic as adoption outruns them, and win public support for data centers — because demand already exceeds supply.
~14 minLip-Bu Tan on the CPU Renaissance
Intel's CEO argues the AI buildout is permanent but mispriced in the short run: compute is shifting from a GPU monoculture toward a CPU-GPU-accelerator mix, agents are the new demand driver, and discipline plus foundry trust decide who survives the coming correction.
~40 minAvsec & Gayoso on the 98% of DNA we called junk
DeepMind's AlphaGenome reads the 98% of DNA once called junk, and AlphaGenome 2 now resolves it cell type by cell type, to rank which single mutations drive disease, from a rare-epilepsy case to a gene regulated 290,000 bases away.
~11 minSarah Friar on Training AI at Altitude
OpenAI's CFO reads a $50B-ARR quarter as the payoff of a deliberate sequence: land enterprises on cyber, train new products on their hardest problems first, then ride the frontier down the cost curve toward 1.2 billion consumers.
~90 minVijoy Pandey on why two AI agents still cannot book you dinner
Six months after multi-agent systems went into production, Cisco's Vijoy Pandey open-sources the protocols that let agents align and negotiate, and argues the hardest unsolved problem is still getting two personal agents to simply agree on dinner.
~86 minZeno Rocha on why agents need your API, not your app
The Resend founder on reinventing email deliverability by trusting senders by default, going down to the metal on infrastructure, and making the whole product formless — API-first, MCP-covered, and readable by agents that can't run your JavaScript.
~57 minStefano Corazza on why creative AI is a loop, not a one-shot
Canva's head of AI research argues that useful creative AI comes from a fast, controllable loop rather than one-shot generation: own your data and train smaller models, separate layout from pixels, judge with rubric-driven VLMs plus human review, and keep driving inference cost down because the field is far from its plateau.