archive
REX·DAILY
Friday August 14 Anno 2026 · Ibaraki, Osaka
I Daily Wisdom
Robert Greene The Daily Laws August 14 · Leave People with a Feeling

For most of us, the conclusion of anything—a project, a campaign, an attempt at persuasion—represents a kind of wall: our work is done, and it is time to tally our gains and losses and move on. Lyndon Johnson looked at the world much differently: an ending was not like a wall but more like a door, leading to the next phase or battle. What mattered to him was not gaining a victory but where it left him, how it opened onto the next round. He kept his eye on the future, and on the kind of success that would keep him moving forward. Johnson used the same approach in his efforts to win over voters. Instead of trying to persuade people to support him with speeches and fancy words (he was not a good orator anyway), he focused on the feeling he left people with. He knew that persuasion is ultimately a process of the emotions: words can sound nice, but if a politician leaves people suspecting him of being insincere, of merely plugging for votes, they will close off to him and forget him. So Johnson worked to establish an emotional connection with voters, and he would close his conversations with them with a hearty handshake and with a look in his eye, a tremor in his voice, that sealed the bond between them. He left them feeling that they would see him again, and he stirred emotions that would erase any suspicion he might be insincere. The end of the conversation was in fact a kind of beginning, for it stayed in their minds and translated into votes.

Daily Law: Keep your eyes on the aftermath of any encounter. Think more of the feeling you leave people with—a feeling that might translate into a desire to see more of you.

The 33 Strategies of War

, Strategy 22: Know How to End Things—The Exit Strategy

Steven Pressfield The Daily Pressfield Day 226 · GENRE GETS A BAD RAP I

emember when Above the Law got reviewed in Daily Variety. The paper gave the film’s title and then, right after that, wrote “Exploitation.” In other words, that was the genre Variety had assigned to it. I was outraged. “How about ‘cop thriller’? What’s wrong with ‘detective story’? I’ll even vote for ‘redemption saga’!” Unpublished Black Irish JAB, “Go Back to Genre” To call a movie a “genre film” is an insult. The term implies a slavish adherence to formula, devoid of originality or creativity. But every book is a genre book, and every motion picture is a genre film. What’s wrong with that? Romeo and Juliet is a love story. That’s a genre. The Odyssey is a hero’s journey saga. That’s a genre. Even the Bible falls into a genre, and if you don’t believe me, ask Cecil B. D eMille.

Leo Tolstoy A Calendar of Wisdom August 14

People have become too accustomed to the use of force.

Life without violence seems impossible for them to imagine.

Two ways exist to guide human activity.

One is to force a person to act against his wishes; the other is to guide a person’s wishes, to persuade him with reasoning.

One is the way of violence: it is used by ignorant people, and it leads to complete disappointment.

The other is supported by experience, and is always successful.

—ABRAHAM COMB

Without faith in God, you can force people to act, but you cannot convince them.

You can become a tyrant, but not a teacher.

—GIUSEPPE MAZZINI

People are wise beings; they possess the ability to live according to the dictates of their intellect, and sooner or later, they will evolve from a state of violence to a state of complete harmony and understanding.

And every act of violence makes this time more distant from today.

II The Day Ahead
  • Tomorrow · 10:50 SQ check in opens: grab free seats now (outbound, all 3) ewijaya@gmail.com
III Signals & Dispatches

AI / ML

  • @OpenAIsource ↗ OpenAI is previewing GPT 5.6 Sol ultrafast mode at up to 14 times the speed, initially through the API for selected customers.
  • @reiraxbtsource ↗ A new digest points to Andrej Karpathy's free two hour lecture as a practical route into building AI workflows.

Drug Discovery

  • @ericxingsource ↗ A new Nature Medicine perspective proposes AI driven digital organisms as dynamic, multiscale simulators for disease and drug development.
  • @defileosource ↗ A new scientific agent skills project claims to connect AI agents with genomics, drug discovery, clinical trials, PubChem, ChEMBL, and docking workflows.

Biotech

  • @__dwr__source ↗ A useful biotech AI constraint is feedback time: coding agents get rapid compiler feedback, while CRISPR design agents face much slower experimental validation.
  • @NuryVittachisource ↗ A US judge temporarily blocked a WuXi AppTec ban after finding serious ownership math errors in the government's case, with wider implications for biotech trade restrictions.

Tech

  • @techreviewjpsource ↗ MIT Technology Review Japan highlighted a report on the shift from data centric AI toward agents that could reshape scientific research.
  • @stevenstrogatzsource ↗ Steven Strogatz shared a free lecture series on asymptotics and perturbation theory for anyone wanting a stronger mathematical foundation.

Japanese Politics

  • @takaichi_sanaesource ↗ Sanae Takaichi strongly protested Vladimir Putin's reported visit to Etorofu and reiterated Japan's claim over the Northern Territories.
  • @kokkai_clipsource ↗ Four newly released Diet committee records include debate on flag damage penalties and amendments to Japan's industrial technology law.
IV Sky over Ibaraki
☀️ +27°C
humidity 66% · wind ←5km/h
clear morning with a light breeze
V What Rex Can Do Today
  1. Review the latest AI drug discovery signals and draft a one page note on where the real bottlenecks remain after hit finding.
  2. Prepare a compact comparison of local Qwen 3.5 27B inference options for keeping proprietary peptide work on premises.
  3. Triage recent work threads and turn them into a short action list for the rest of the week.
VI Toward the Long Game
  1. Read one primary paper on AI assisted molecular design and write five sentences on whether its validation is genuinely translational.
  2. Run a small reproducible benchmark comparing a protein or peptide representation model with a simple classical baseline.
  3. Spend thirty minutes mapping one Japanese biotech or AI drug discovery group worth following, then identify a concrete networking route.