Section path enables citations and whole-section retrieval
The section path stored on each chunk is used both to render citations and to fetch an entire section in order (sorted by chunk index) when the agent needs more context.
37,436 results · Most relevant
The section path stored on each chunk is used both to render citations and to fetch an entire section in order (sorted by chunk index) when the agent needs more context.
Dense embeddings capture meaning but treat rulebook codes and exact identifiers as noise, while BM25 handles exact tokens but not paraphrases; the team combines both in one query call, fuses with reciprocal rank fusion (dense top 50, BM25 top 50, fused to top 30, re-ranked for section coherence, top 10 to the agent), and notes RRF needs no score calibration between legs.
A lot of the important knowledge is structured within the rule books' ~3,000 tables, which the team initially did not realize. The initial table processing just fetched plain text from the table in the PDF, which did not format tables well: if table cell spans are not clearly represented in extracted text, the agent can fetch wrong information — especially dangerous with important things like fees. In a dummy example, the domestic field spans multiple rows, cross-border over two more rows, and the refund spans over both. The fix: transform tables into CSV format, duplicating spanned fields so the agent fetches the correct row it needs, which made a big difference for the correctness of the answers.
The core serving fix: pre-compute embeddings only for users and merchants, never per transaction, cutting the cache footprint from 60 GB to 2–3 GB — about 20× smaller and cheap enough to run on the same CPU class as the XGBoost model.
The host says you need friction to let people into your dating life, your company, and your friend circle, and that this friction can simply be asking people what they're about without tipping your hand. He advises letting people show up, show you their track record, and tell you what they're about instead of pitching and overselling. He warns that overselling makes people get excited just because you seem like you've got a mission and they feel like they could use one, and that getting so good at talking people into things that they enthusiastically assent in the moment but regret later creates more problems than you've solved. He describes his own former 'inspiring halftime speech' pitch — 'Here's what we're doing. It's going to be amazing. You should be part of it.' — and says his fake,
Prioritize selling brokerages over individual realtors: go to the head of brokerages (which have 5, 10, 20 agents) and offer a brokerage-level deal — free VIP upgrades for all their agents if you can be the first people they call for a move. One brokerage deal is worth roughly 60 houses a year; 20 agents each selling about a house a quarter equals about five houses a month, which at a $3,000 average is $16,000 a month per brokerage. Bring on a new brokerage every week, and each adds roughly five deals a month as it stacks. Brokerages are the long-term play where reviews compound.
Future projections — suggesting you and she will do something together later — are rated S-tier. Introduce them as a hypothetical joke (e.g., 'are you going to travel with me or what?'), then dial back to a smaller real plan like grabbing coffee first.
The coach strongly advises against telling a woman you're approaching that you go around approaching women. He explains it makes the interaction a lot less unique, is a huge red flag, gives off a 'fuck boy' vibe, and for many men will be the end point of the conversation — for some women an instant deal breaker.
Before trusting someone, verify two conditions. First, do they gain more from protecting you than from betraying you — does burning you cost them more than protecting you? If betraying you would be a good deal for them, the incentive to wrong you is too big, and you shouldn't share something material with them. Second, do they have a track record of protecting what they've been given — have they had a knife pointed at your back and chosen not to use it? Never trust someone who has already burned you, and never trust someone who hasn't followed the prescription of becoming trustworthy. Only when both conditions hold does it make sense to go forward with the trust.
Consequential creator advice requires independent verification before Today can promote it.
The fourth and most robust strategy is to use Kubernetes or another highly scalable ephemeral environment. You get a pod for every agent session: it spins up, the agent runs and does its work — maybe on a local copy of the repo — then spins down and the environment goes away, with the agent having opened a pull request. You don't have to worry about the long-term state of the environment the way you do with a VM, because every time an agent runs you're killing the environment and starting fresh. Kubernetes gives you infinite, cloud scale — you can run as many agents as you want as long as you're willing to pay for the memory and CPU.
Hormozi instructs the audience to first calculate the true costs of delivering their service at different levels, then set prices at 5 to 10 times that cost. He warns that without knowing true costs, you cannot apply this multiple and will keep making up numbers, and he notes that everyone in an industry marks up prices — most just aren't organized about it.
Hormozi says never to present an invoice or proposal via email — always over the phone. Presenting live lets you resolve concerns and answer questions in real time, which is what gets clients to buy. Joey corroborates from experience: when his company sent proposals via email, clients replied with objections like 'that feels like a lot' or 'my needs have changed' that couldn't be addressed in real time, and deals were lost.
Hormozi recommends building a dynamic pricing sheet — coded with AI or as an if-then Excel sheet — that takes event headcount as the primary input, derives variable costs proportional to headcount, adds fixed costs, and outputs the final price automatically. This standardizes pricing so sales reps can quote without long delays, and without it, teams make up numbers on line items and send inconsistent proposals. He notes such a tool is a programmable AI task doable in a couple of hours.
Use outcomes to create self-supervision: you tell Claude the task and a rubric to evaluate the task, and it goes off and does the work. When an outcome is set, a sub-agent looks at the rubric and the work Claude did and evaluates whether it did it correctly and to completion; if not, it tries again. A user-defined outcome event is implemented so that any time a session is started in the UI for a specific ticker, it sends the outcome with a specific rubric, max iterations, and a description of the task.
Use outcomes to create self-supervision: tell Claude the task and a rubric to evaluate the task, then a sub-agent evaluates whether the work was done correctly and to completion — if not, it tries again.
Configuration hardening is the fourth layer of AI security, covering the AI-native IDEs and CLIs themselves (Claude Code, Cursor, Windsurf, etc.). Keep them updated to the latest version, though it can help to wait a few hours or one day after a new IDE version is released since updates often fix issues or memory leaks. Minimize the command allow list to only what you need, disable auto-approved commands wherever possible (especially for disk operations), configure the tools to use private models if available to minimize data-poisoning risk from a downloaded model, restrict them to specific project directories only, and monitor for unexpected IDE behavior.
Maintain an inventory of all components (AI BOM) that make up agentic applications to prevent supply chain attacks. When a zero-day is disclosed for a specific model, dataset, or MCP tool, use the AI BOM inventory to quickly identify everything impacted in your organization and fix it.
Define trust boundaries in AI systems — data injection or content construction points where external input enters. Sanitize input from APIs (e.g., weather.com or a stock exchange database) because you shouldn't trust it; it could be an indirect attack. External data influence and downstream actions, when AI outputs trigger system actions, are also trust boundary concerns. Malicious content planted on LinkedIn could be processed by LLMs and lead to leaked data.
When defining an agent, start by setting its boundaries, including what error tolerance is accepted. Error tolerance targets vary by context: someone might want zero or 90%, but sometimes 50% is good as well, since AI is not always perfect. In autonomous environments with no human in the loop, agents should have a higher accuracy target — it should be perfect or the job should be cancelled. In human-to-AI collaboration where AI creates content and a human makes the final decision, a 60% error tolerance might be good.
The intern metaphor breaks when comparing hiring a person versus hiring an agent: when you hire a person, the interview ends, but when you hire an agent, it does not. Unlike a person, an AI agent's evaluation never ends — you rerun it every week on live traffic against a human baseline: instrument, evaluate, improve, retest. A dashboard can show a flat 92% containment score all year while the real measured score slides from 83 to 60 — same agent, same green dashboard, but different metrics over time. Weekly measurement of deployed agents can reveal real performance decline that a static dashboard hides.
Choose sources → to get a library of your own and a daily brief.