Use agents to create and maintain Code Connect documentation
Jake recommends using agents to create and maintain Code Connect documentation over time, rather than doing it manually, to keep it up to date as designs and code change.
35,595 results · Most relevant
Jake recommends using agents to create and maintain Code Connect documentation over time, rather than doing it manually, to keep it up to date as designs and code change.
The host's core advice is to sell your SaaS to whomever pays the most at its most valuable moment, and not to mess it up.
Tara Seshan observes that the most effective people use AI tools to expand the set of things they are capable of doing, not simply to automate rote tasks. They expand their range of possibilities so they can realize more of what's in their head into reality. Everyone now has the superpower to spin up designs, build prototypes, and model scenarios.
Richard Socher recommends regulating specific applications of AI rather than intelligence in the abstract, and warns against limiting compute (flops, GPU tracking) as that would require a totalitarian state.
Rob Walling advises that after a day or two of consuming content, you should pick up the phone and make your first cold call. After the call, go back to resources to improve: check a subreddit, ask a coach, record the calls for feedback, or ask a peer mastermind group for advice.
Rob Walling recommends learning to regulate your nervous system yourself, potentially with help from a therapist or medication. Specific techniques include slow breathing, self-talk (e.g., 'it's going to be alright, this is not reality right now'), and playing a handful of songs on loop that pull you back into a place of calm.
Consequential creator advice requires independent verification before Today can promote it.
Rob Walling outlines a process for learning cold calling: first, consume content from trusted sources like Steli FD, close.com, and TinySeed mentors for a day or two to get a frame of reference. He recommends consuming all the content in a few hours, in the background while doing dishes or for an hour or two during your work day.
Consequential creator advice requires independent verification before Today can promote it.
Derrick Reimer and Rob Walling advise founders to overcome social anxiety by forcing themselves to get reps in (e.g., sales calls, public speaking), as confidence builds with experience. They also recommend recording sales calls and feeding the transcript into an LLM for objective feedback on performance.
confident
Gonen Stein and Ofir Ehrlich recommend that organizations map, classify, and tag all their data resources to ensure they are protected. They cite a case where a customer lost 60% of their environment to ransomware because they failed to do this, despite thinking they were protected.
Clay Parker Jones recommends turning psychological safety from a platitude into a structural thing by writing down the rules of the road, measures, and how direction is set. This shifts the focus from interpersonal relationships to structural rules about business outcomes, making it arguable on a basis of fact rather than opinion. The goal is for people to know why they are doing something and what is expected of them.
Jeff Cavaliere recommends the cheek curl for eccentric overload on biceps. He instructs the audience to allow a little momentum, lean to vertical, lock that position, then slow the weight on the way down. He advises not going all the way back through vertical. The audience should drop the weight down very slow to a count of three, cheat it back up to the top, and perform a set of eight reps to failure, training to eccentric failure where they can't even control the weight anymore. On day 13, they should increase the eccentric lowering to four seconds per rep, doing as many reps as possible under that control. On day 20, they should start with a five-second eccentric, then drop to four-second eccentrics for as many reps as possible, then three-second eccentrics for as many as possible, and
Jeff Cavaliere recommends the PJR pullover for triceps eccentric overload. He instructs the audience to get into a position similar to a dumbbell pullover, start with the dumbbell at the top, and lower it down slowly for eccentric overload on the triceps. He recommends using a heavy weight in the 8 to 10 rep range and using the lats to pull the dumbbell back to the top, as the concentric phase is not as important as the eccentric lowering. The audience should shift the focus back to the triceps, particularly the triceps long head, as they lower the dumbbell down slowly. They should start with three-second eccentrics in week one, increase to four-second eccentrics in week two, and on day 21, raise the eccentric to five seconds, then drop to four, then three until they can't control it a
Jeff Cavaliere recommends stepping backward instead of forward when lunging to reduce stress on the knee joint.
Ramakrishnan Lokanathan's rule: when your eval scores reach 85-90, that is the time to change the test and make it harder. He uses the SAT analogy: if everyone is scoring in the 99th percentile, the test itself has to change.
Charlie Holtz advises being quite careful with certain parts of the codebase while being very loose with other parts, rather than pure token maxing. He cautions against assuming high-output AI teams are ripping through 30,000-line PRs without review; they are actually careful with critical code. He also notes from experience that not being careful about slot-free zones forced the Conductor team to rewrite their whole app a couple of times.
A high-level intelligence-plane architecture is described: an agent runtime executes the agent, with memory and knowledge (a memory manager) sitting on top of a data platform. The agent runtime writes traces of memory into the memory manager, and memory traces feed into a continual learning loop that changes models back into the agent runtime.
hedged
A convincing image can hide how much of an idea is still unsolved, and a prompt like 'fill a room with LEDs' answers almost none of the questions about how light creates experience, depth, or how movement feels around your body — so you have to make all the decisions down to the details yourself.
The inference gateway provides resilience: if one provider goes down, Optiver can swap to a different provider, because any time they are down they are not making money.
Optiver benchmarks new models on golden sets collected from key customer trading teams to determine if they are valuable; this requires a mature evaluation framework and partnering with those teams to collect their benchmarks.
If you do OCR on a figure and want specific data from it, you lose all the information and cannot match the page.
Choose sources → to get a library of your own and a daily brief.