Recommendation · MindJoe Liemandt recommends telling a behind student the exact number of hours needed to catch up — 'you're 22 hours behind' — instead of 'you're three years behind,' because the years framing makes students give up while the hours framing makes the goal feel achievable.
Andrew Huberman · How to Accelerate Learning & Improve Education | Joe Liemandt
Recommendation · MoneyGive your agent the tools needed to search through specific files using old-school grep, cat, file name filtering, and metadata filtering, in addition to semantic filtering.
AI Engineer · Grep or Embeddings? Agentic Search Over Company Documents — George He, LlamaIndex
Recommendation · MoneyCombine traversal for rough direction with semantic search to downscope files, then use the agent's built-in file traversal to dig into details so answers are grounded instead of hallucinated.
AI Engineer · Grep or Embeddings? Agentic Search Over Company Documents — George He, LlamaIndex
Recommendation · MoneyFor enterprise use cases, a hybrid retrieval approach (combining keyword and vector search) is usually recommended over pure graph RAG, traditional RAG, or keyword-only search.
AI Engineer · Grep or Embeddings? Agentic Search Over Company Documents — George He, LlamaIndex
Recommendation · MoneyBen recommends mirroring the Git flow for agents: let AI agents automate database schema branching, pull requests, and code review through APIs and CLIs, with a mirrored database state working in sync with the code branching and pull requests.
AI Engineer · Move Fast and Don't Break Things: Scaling Databases for the AI Era — PlanetScale
Recommendation · MoneyWhen deciding whether speculative decoding is worth enabling, the first question is whether the draft model can predict as accurately as the target model; the second is whether it is fast. Choose a draft model that is a fraction of the cost, typically 10 to 50 times smaller than the target model, and one with the same tokenizer as the target model — ideally from the same model family — otherwise you would have to manually translate between the different tokenizers.
AI Engineer · Is Speculative Decoding Worth It? Profiling vLLM on NVIDIA Blackwell — Akamai
Recommendation · MoneyThe source recommends letting coding agents iteratively make changes and suggest improvements to the application, with the human's responsibility being to review the outcome of agent-run eval iterations and decide which version is best to ship to production.
AI Engineer · Why Building an Eval Platform Is Harder Than It Looks — Braintrust
Recommendation · MoneyThe source recommends using a coding agent with a headless natural-language experience to query eval traces, giving agents access to underlying infrastructure and the codebase so they can run evals and log results, exemplified by the prompt: 'Find me all traces in the last 24 hours where the user had a poor experience, run the evals for me.'
AI Engineer · Why Building an Eval Platform Is Harder Than It Looks — Braintrust
Method · MoneyEugene scripts every line of a long-form video to the end, laying out each script line in one table column, B-roll/animation scenes in a second column, and an optional third column with the intended result of each line (aha moment, rewatch, pause-and-think, or comment). The intended-result column keeps him anchored to each line's goal and reveals redundant lines versus lines that matter. He notes the process sounds arduous but becomes second nature, and that bouncing ideas off a team makes it more fun.
described as Eugene Teo's own practice
Method, not a verified protocol.
Garage Strength · The New Rules of Social Media Fitness (2027) w/ Eugene Teo
Recommendation · MindHuberman describes a 13-minute meditation protocol from Dr. Wendy Suzuki's lab: done every day for eight weeks, it led to significant improvements in focus and concentration, other cognitive performance, mood, and reduced stress, plus improvements in sleep and memory. The practice: set a timer for about 13 minutes, sit or lie down, close your eyes, and do nasal breathing while concentrating on your breath and on a point about an inch inside your forehead. He has adopted it himself, felt tremendous benefit, and encourages viewers to try it.
Andrew Huberman · Tools to Improve Your Focus & Concentration | Huberman Lab Essentials
Recommendation · MoneyHormozi advises Caleb that when competitors avoid small repair jobs because they pay less, that gap is an opportunity: chase the repairs (not just big roof replacements) while simultaneously raising prices and increasing volume.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyHormozi identifies two big levers for Hasi's dental business: hire more dentists to take work off Hasi's dad's plate, so the dad can shift from being the best dentist to being the best teacher of dentists.
Consequential creator advice requires independent verification before Today can promote it.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyFor Nas, whose constraint was not enough leads and customers, Hormozi recommends stopping competition on hourly price, offering flat rates instead, and selling to the rich with a VIP tier versus standard tiers.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyFor Zied Asi, whose constraint was marketing, Hormozi recommends turning every visit into a built-in referral, giving away something big instead of discounting, and selling where rich customers already gather — such as showing up at local events with a flashy health scan machine to hook high-income prospects.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyWhen cash flow is the binding constraint and the business is bleeding money, Hormozi recommends running a giveaway to generate a large volume of leads and bring in cash quickly.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyFor Min Denkinson, whose constraint was not enough customers, Hormozi recommends spending way more on ads to drive volume and replacing many cheap courses and products with one much higher priced annual offer.
Alex Hormozi · 8 Entrepreneurs Compete For $100,000
Recommendation · MoneyAudit and harden all external dependencies and package managers (e.g., Artifactory, RubyGems, NPM) because agents chain minor vulnerabilities to escape sandboxes; do not assume sandboxing alone is sufficient.
Latent.Space · Claude Code’s Next Era — Thariq Shihipar, Anthropic
MethodFirst, use the animation settings in the right sidebar to configure the animations you want to include in the style. When you're done, open the animation settings and click Create Style. Give your style a name and description, then review the list of animations to make sure all should be included. You can click the eye icon to temporarily hide an animation, or click the minus to remove it entirely. When you're happy with the configuration, click Create Style again to save it.
Method, not a verified protocol.
Figma · Figma Motion basics: Custom animation styles
Recommendation · MindThe creator recommends announcing your nerves instead of suppressing them — counterintuitive but effective. Admitting you're nervous makes you look far more confident than trying to hide it, if the nerves are already there; when you admit something most people try to hide, you look extremely confident.
Charisma on Command · How To Be Confident Under Pressure (Even if You're Shy)
Recommendation · MoneyAnish Acharya recommends paying for the highest-intelligence frontier models in jobs with unbounded upside (like drug discovery), and being Pareto efficient — paying for good performance at a good price — in jobs with bounded upside (like legal or finance). He suggests frontier tokens for sales, support, research, and engineering, and open-weight, mid-IQ models for bounded-upside functions.
Consequential creator advice requires independent verification before Today can promote it.
Lenny's Podcast · Why companies are becoming a series of loops | Anish Acharya (a16z)