kern

Investing in what’s next, roasting what’s dumb🤗

[kern @kernbz]

@rayruizhiliao Good vibe

[kern @kernbz]

@jjschnyder Nice vibe in Zurich!

[kern @kernbz]

@julien_c For sure

[kern @kernbz]

@PariLatawa Need beer 🍺

[kern @kernbz]

@Kalshi Why should we care about them?

[kern @kernbz]

A corporation can survive everyone who remembers why it was founded. This is one reason its purpose must occasionally be rediscovered in a workshop.

Organizations attain a lifespan that transcends the individual, and with it, the capacity to forget their own origins. Missions do not perpetuate themselves naturally���slogans, however, do.

[kern @kernbz]

@NanouuSymeon Hah, i am good, morning!

[kern @kernbz]

@kaifulee I like the opening where you describe how, at the time, you viewed AI as “the final step in humanity's understanding of itself.”

[kern @kernbz]

@anah_sahh innovative tech needs trust, Apple dose!

[kern @kernbz]

In a market, people do not merely predict events. They predict how other people will predict events, which is how a mirror eventually develops monetary policy.

[kern @kernbz]

Wealth is not freedom from consequences so much as the ability to let consequences travel through several other people before arriving.

Inconvenience, risk, waiting, pollution, and exhaustion have not disappeared; they have simply acquired longer transport routes.

[kern @kernbz]

@tonyzzhao I like the hat design

[kern @kernbz]

AI can produce ten confident answers in the time it takes doubt to locate one reliable source. Confidence has finally achieved industrial scale.

In the future, the scarcest resource may not be answers, but rather those willing to delay them. Validation remains a slow process, for reality lacks a "generate" button.

[kern @kernbz]

Wow, this is Crazy!

[Haven Feng @HavenFeng]

Today, we’re introducing [schema]: a harness reaching 99% RHAE with Opus 4.8 + Fable 5 and 95.35% with GPT-5.6 Sol on ARC-AGI-3 Public set.

[schema] makes an LLM think like a physicist. 🧵

[Haven Feng @HavenFeng]

ARC-AGI-3 gives an agent a 64×64 grid plus legal actions, no rules, stated goal, or reward. The agent must discover both what the world is and how it works like a physicist:

1. State grounding -> identify objects, relations, and goals.
2. Mechanism discovery -> infer how these states change.

[Haven Feng @HavenFeng]

Today, we’re introducing [schema]: a harness reaching 99% RHAE with Opus 4.8 + Fable 5 and 95.35% with GPT-5.6 Sol on ARC-AGI-3 Public set.

[schema] makes an LLM think like a physicist. 🧵

[kern @kernbz]

@naval This is crazy!

[kern @kernbz]

@HavenFeng This is crazy!

[kern @kernbz]

what will happen in education ?

[kern @kernbz]

I hope to see more robot companies competing fiercely🔥

[Bernt Bornich @BerntBornich]

The NEO's in the lab keep talking to each other