Browse entries 1513–1536 of 1597. Return to the first page to search and filter the complete collection.
Play video
Getting an AI agent to behave the way you want isn't just about writing better prompts. In real systems, behavior emerges from a loop: prompts, evals, iteration, and feedback. Small changes in any part of that loop can completely change outcomes.
Play video
An agent runs entirely on your phone, no cloud, and plays Space Invaders, perceiving the scene, predicting the aliens, and dodging bullets in a loop. Another solves the New York Times mini crossword with a constraint graph that backtracks when the fills stop fitting.
Play video
An agent hands a doctor a clean, confident fact: the patient has a penicillin allergy. But that fact was synthesized from three sources, an EHR record, a lab report, and something the patient typed into an intake chatbot, and by the time it reaches the doctor, which one it came from is gone.
Play video
In July 2025 Dex Horthy turned the lights off: an agent software factory where nobody read the code. It fell apart. An issue appeared that no amount of prompting could fix, the site was down, users were furious, and he was digging through a codebase he had stopped reading three months earlier.
Play video
A second refund on the same order. A payout sent to the support desk instead of the buyer. An order status of "probably shipped." These are the kinds of mistakes a probabilistic agent makes and a paragraph of instructions cannot reliably stop.
Play video
Your agent can reach your data and still get it wrong. Vector search hands it a slice, Text2SQL hands it another, and neither tells it what is actually relevant or how the pieces connect, so the answer comes back confident and wrong.
Play video
Feed it 67 videos from the 2022 World Cup and ask for the near misses, the shots that almost scored but did not, each with a reason, and it returns them. Ask it to track Messi across the entire corpus and describe the camera framing, and it finds the moment he slaloms past a sliding defender.
Play video
Inference traffic at Meta already outpaces the largest microservices in the world, the fastest growing workload it has ever run. Nishant Gupta's frame for what comes next is 2008.
Play video
A single mid range GPU can turn half a million tokens per second into embeddings in the low tens of milliseconds, where a managed endpoint costs orders of magnitude more and takes hundreds. Daniel Svonava calls embeddings the no brainer entry point. His real subject is what comes after.
Play video
Ask a benchmark harness for 200 queries a second and it may quietly deliver 38, then print results as though it ran 200. Ashok Chandrasekar opens with that experiment, which is why he and Jason Kramberger, both at Google, kept failing to reproduce published numbers.
Play video
Split a job across several nodes, let their GPUs compute, then have them exchange a little metadata before the next round. While that exchange happens every GPU sits idle, and the whole cluster waits on the slowest one. Back when a compute phase ran five seconds and the exchange took a few milliseconds, nobody cared.
Play video
People used to type "am I fat?" into Yahoo Answers. Edo Liberty worked there at the time, and the question stuck with him, because a Q and A forum has no possible way to know.
Play video
Legora's search latency once went from a 100 millisecond P99 to twenty seconds, and the cause was a packing problem invisible in the schema.
Play video
Ask your own company what it actually contracted for with a given vendor and nobody can tell you. The number exists, negotiated carefully, sitting in a pricing table inside a signed PDF no system ever read back.
Play video
Ask a corpus of Seinfeld transcripts for the name of Jerry's favorite church and a 100 token chunk returns it at rank one, while every larger window buries it below rank fifty.
Play video
Asked in Spanish about a mid century sofa, the model answers in Spanish but leaves the phrase mid century in English, because that is how the term is actually used by Spanish speakers. Nobody wrote a rule for that.
Play video
Amit Desai never makes the model better. Accuracy stays pinned at 79 percent for the whole talk, and user pain still drops by nearly half.
Play video
Tolan's response latency once drifted from two seconds to about two and a half. That half second tanked essentially every metric in the product, and users wrote in to say their companion had become slow.
Play video
Taste Labs analyzed more than two million websites going back a decade, and the finding that complicates the usual story is that the internet was homogenizing before AI arrived: palettes converging, layouts converging, trends spreading faster than they once did. AI did not start the collapse.
Play video
The instruction Impeccable loads when you type "bolder" ends with a challenge: show someone your work, say AI made this bolder, and if they believe you, you failed. Bakaus has built tools for more than twenty years.
Play video
Why porting a massive codebase from Python to TypeScript was the right move. Mike Krieger, co-lead of Anthropic and founder of Instagram, breaks down the unconventional engineering decision behind this switch.
Play video
Ask a model to sum twelve numbers scattered across thirty thousand tokens and it will often get it wrong. Give it a Python REPL and it writes a few lines of regex instead. That is the shift Kevin Madura, director of advanced technology at AlixPartners, describes in recursive language models.
Play video
Tanmai Gopal plotted the daily edits to his own company brain expecting the usual shape, a burst of enthusiasm followed by neglect. The line kept climbing instead, and it surprised him.
Play video
Gus Iwanaga opens by disowning his own session title, then shows the demo that did not work. His team asked their system for a sales report for Q1 four times over.