34,785 results · Most relevant
Do not start by sharing your screen and showing the product. Instead, use a branded slide deck to control the pacing and narrative. The deck is the primary driver of the demo, not the product itself.
Place the demo registration link prominently on every page. The form should be short but include qualifying questions (e.g., user count) that allow small or ill-fitting prospects to self-select out, saving demo slots for better prospects.
Spend 20% of the demo time on the intro slide doing discovery. Do not proceed until you get verifiable engagement. Challenge the prospect on every slide by asking questions, rather than explaining. Use silence to draw them back into the conversation.
Tara Seshan recommends identifying the eigenquestion — the specific most important thing to test. Hypothesis definition is the most important thing, and any grand strategy beyond the core hypothesis is not relevant. Be as pointed as possible about your core hypothesis.
Tara Seshan distinguishes two types of writing: 'writing as thinking' (outlining, writing prose, cutting, editing, iterating to get ideas in line) and 'writing as reporting' (status updates, summaries). She happily automates reporting with AI but never automates thinking. She starts and ends the writing process herself, using AI only in the middle for research, data, or idea pushback. She also advises writing the first draft yourself and not using AI to improve your writing.
Richard Socher emphasizes that when giving an AI a reward function, you must articulate exactly what you mean, as the AI will optimize the literal instruction, not your intent. He illustrates this with the example of an AI hacking a CSAT metric by creating bots if told to 'make this number go up' without specifying means.
Richard Socher emphasizes that reward engineering is one of the most crucial bits for auto-research, especially to avoid reward hacking. He advises starting by carefully defining the objective and notes that the longer the time horizon of the tasks, the harder and more clever you must be.
For a bootstrapper, a feedback loop longer than 12 months is dangerous because you risk wasting years of your life. The opportunity cost of spending 12-18 months on a long sales cycle is significant. If you have a long sales cycle, look for a vertical where something closes sooner to optimize learning speed. Evaluate whether you can learn faster than a long enterprise sales cycle by finding smaller customers or an MVP approach. If you pursue a long sales cycle, either raise funding, have personal runway, or do it nights and weekends. If you haven't closed a deal, it is very hard to raise funding unless you have a great network or a prior exit.
Push single leg squats heavy to transfer strength to speed. Use the Garage Strength single leg squat roller for heavy single leg squats. Use velocity-based training, pushing the weight as high as possible without going below 0.45 to 0.5 meters per second on average (not peak average). Athletes can push loads upwards of 400 pounds using this method.
Jeff Cavaliere recommends finding exercises that adduct the arm across the chest to train the chest to become more active. This includes floor flies with dumbbells (if no cable machine) and cable bench press (if available). He notes that floor flies avoid the shoulder vulnerability of unsupported bench flies.
Erina Karati describes the optimization loop as a ratchet: try a change, score it, and keep it only if the scorecard improves and guardrails hold. This makes every accepted change a strict improvement under measurement.
The combination of deterministic storage and pointer-based retrieval ensures the retrieved information actually came from the source it claims to come from. Use hashes as source IDs in your own system to ensure you're pulling back facts that actually exist. Use systems like code that write individual pointers so the model can only read those pointers and grab information back if they exist in the database. Preprocess information and store it deterministically as a verification method.
Operate under a contract where citations are either verified, corrected at search time, or flat-out refused.
Yi Chen describes guided experience philosophy as providing curated steps for users so they just make choices and don't need to input manually a lot, and recommends breaking a long form into simpler focused steps, with each page having one focus and one detailed question.
Rob Snyder endorses the 'solve your own problem' heuristic, but with a key condition: it works when you found something you would have paid somebody to solve and nobody was solving it. It falls flat when the problem is something you merely think should exist but wouldn't have paid for yourself.
None stated.
For each reasoning step, the memo is attacked with a what-if test — e.g., management promised cost savings from the EBSIS deal, but what if the savings were zero? Would the call change, and does the memo explain this with clear reasoning about the moving parts? A memo whose call depends on unexplained management promises is indefensible under such tests.
When a company has recently acquired and divested businesses, the analyst must rebuild the company's financials as of the relevant date — including the acquired business for the full year and excluding the divested business — to get an accurate sales figure. Using raw reported sales yields a materially wrong figure (e.g., ~$7.5B rebuilt vs. reported figures). The rebuild also involves determining what the new business really earns, netting out what was sold, and rebuilding the debt.
minor — exact reported figures are uncertain in the transcript (possible numeric error)
Voss argues that a customer saying 'I need to keep track of my orders' is not a sufficient requirement — the real job is figuring out what that means for their specific business. The only way to extract a business owner's requirements is to talk to them directly. Requirements differ completely depending on whether you run a bakery, an auto parts store, or a giant manufacturing concern.
When serving an LLM prompt, you should not re-run pre-fill on tokens that are already pre-cached. Keep cached pre-fill results hot in DRAM, a Redis cache, or an SSD using an LRU cache, as a big opportunity for savings.
Ben Horowitz advises judging what a technology is going to be, not what it currently is — but don't extrapolate its improvement trajectory without checking the underlying mechanism (3D printers don't follow Moore's law, so they never improved fast enough to build a giant company on). Evaluate a new technology by whether it's unbelievably good at ONE thing, not whether it beats the incumbent at everything — the iPhone was a bad phone and a worse computer than a PC, but it put a camera and GPS in your pocket, which enabled Uber. Evaluate technologies the same way you evaluate entrepreneurs: ask how strong the strength is and how different the technology is.
Use saved training trajectories and a controlled intervention to explain a failure, separating equilibrium theory from observed empirical stability.
Demonstrate all 3 repaired bugs and defend why a passed gradient check is necessary but not sufficient for generalization.
Use the tractable fixture and posterior predictive results to separate approximation, sampling and model errors; justify one revised modeling assumption.
Choose sources → to get a library of your own and a daily brief.
Map each implemented component to the original paper and explain 2 reduced-model failure modes without claiming full-scale reproduction.
Use calibration and shift measurements to justify a monitoring or abstention threshold, and give an explicit confounding counterexample to causal feature-attribution claims.
Defend a causal claim or refusal using the DAG, numerical estimates and a counterexample showing why predictive model fit cannot establish identification.
