Joseph

VC at Red River West: investing in founders who make you feel stupid but optimistic
trying to run fast and read far :)

[Joseph @Joe0ussai]

a but of a vibe shift this week: I can talk to my non-AI-pilled friends about Moderna’s cancer news or mainstream companies using “singularity” without sounding like a completely delusional techno-optimist
shared vocabulary and examples could unlock an incredible conversation

[Joseph @Joe0ussai]

24h without codex for the first time since february, AMA

[Ariel Renous @ArielRenous]

Two things Travis does that makes him an exceptional storyteller:

1. He is using metaphors A LOT. Every point he makes during this three minute clip comes with an extremely visual (and simple) metaphor. He's describing VC behaviours talking about chess grandmasters, Michael Jordan and Serengeti antilopes. As a listener, you are hooked because you don't expect this.

2. Repetition. He does not try to rephrase himself. He found a good word: "Everybody wants to make a mark on the world", and he won't let it go. He repeats it like a singer.

[Joseph @Joe0ussai]

@hkanji never been happier or more optimistic on my side!

[Joseph @Joe0ussai]

@gabriel1 Pls list a few examples! :)

[matt rothenberg @mattrothenberg]

this meeting could have been an MCP invoked search of our company’s internal knowledge base that aggregates data from virtually every service we use behind a unified retrieval layer and provides a streamlined interface for extracting critical business insights

[Peter H. Diamandis, MD @PeterDiamandis]

This is the most extraordinary time to be alive. Not the most comfortable. Not the most predictable. The most extraordinary. More tools, more knowledge, more possibilities than any generation before us.

[Joseph @Joe0ussai]

i don’t think we’re talking about the same thing: you’re talking about economic value, i’m talking about the quality of thought
if AI handles my generic scheduling emails and leaves me enough brain to write sci-fi at night, even if nobody pays me for it, that’s still a move toward higher-level, more human thinking
if AI eventually does everything economically valuable better than us, that’s another debate entirely (and perhaps the beginning of abundance)

[Joseph @Joe0ussai]

@SamuelBeek yes I replaced it with agentic coding

[Joseph @Joe0ussai]

@victornunez absolutely, I've stopped counting how many people have told me about business ideas
I reply they should have speech to text'd what they just told me and launched it and then comes the "actually never mind"

[Joseph @Joe0ussai]

i don't have a to-do list anymore. i have unarchived Codex sessions

[Aporia @aporia9n]

I think ambitious people should treat life in “seasons” rather than trying to balance everything every day.

A few years massively overallocated to work. A period getting seriously fit. A year or two nomading. Then family takes priority for a while. Then maybe at 35 you go completely insane again and spend four years building a new company, and so on.

The mistake is thinking every metric needs to stay green all the time. Most great outcomes (building a company, writing sth serious, training for an event) need uninterrupted runs, not daily moderation. Sometimes work should suffer cause you’re travelling, sometimes your social life should suffer cause you’re building, sometimes career progression should slow cause family matters more. That’s fine.

And tbh it probably makes life more fun too. You get actual chapters and completely different versions of yourself instead of spending 40 years maintaining the same perfectly optimised routine. A few intense years in one city, a weird nomad phase, an obsession that takes over your life, then something completely different. Probably much more memorable than one very long well-optimised Tuesday.

It's why I encourage everyone to think of balance in years, not days. A good life can look horribly unbalanced on some random Tuesday and still be very balanced over 10 years.

[Zvi Mowshowitz @TheZvi]

Registering my prediction on this, too: If AI starts replicating Einstein's mental leaps, there will be, by the same people, new and different cope.

[Dylan Field @zoink]

There are three choices. Pick one:

1. AI is conscious, humans are conscious
2. AI is pattern matching, humans are conscious
3. AI is pattern matching, humans are pattern matching

[Benjamin @bschne]

I get the impression AI makes a lot of people addicted to their work the way some get addicted to video games because it lowers willpower/activation energy needed to get to the next incremental bit of reward by taking care of the tedious stuff

[sarah guo @saranormous]

yes, markets/risk management lessons to be learned, but many of the folks dunking on Leopold have never realized a dollar of returns. he has, and now he’s banked them. respect where due

[Will Manidis @WillManidis]

genuinely disgusting the number of VCs sitting on maybe 1x, mostly 0.75x non-mark-to-market, 2021 books spending today tweeting dismissively about one of the most interesting runs in recent asset mgmt history.

try playing in the big leagues buddy, i'm sure you could do it.

[Nicolas Bustamante @nicbstme]

This is basically the first step of the hard takeoff scenario. So it’s not “AI suddenly wakes up and becomes a god”, but AI starts improving the system that runs and trains AI, in domains where progress is measurable and verifiable.

GPT-5.6 Sol is already analyzing production traffic, finding load-balancing inefficiencies, testing new routing strategies, rewriting GPU kernels and validating the results. Those kernel optimizations cut serving costs by 20% (so billions of $$$)

It also designed and ran hundreds of experiments to improve its own speculative decoding model, launched the training runs, monitored them, and intervened when hardware failed or training became unstable. Result: 15%+ more token-generation efficiency (which means freeing more compute and reducing cost).

This is recursive self-improvement in its early, very practical form! The model improves the engineering stack, which makes the model cheaper and faster, which gives researchers more compute and more experiments, which produces a better model…

Today this works best in software engineering because the output is easy to measure: does the kernel compile, is it correct, is latency lower, did cost go down? Soon the same loop moves into AI research itself: experiment design, architecture search, training recipes, data generation, evals, debugging training runs.

At that point it is no longer just “AI helping engineers work faster.” It is AI accelerating the process that creates the next AI.

That is the hard takeoff path to the singularity (AGI). Not one magical jump, but a feedback loop that keeps getting faster. If you want to be fancy you can say: « This is the beginning of a recursive self-improvement loop toward the singularity. »

[Sahil Bloom @SahilBloom]

The ultimate flex in any social situation is just listening to others without needing to talk about yourself. It’s interesting how much it stands out when someone does this. Clear sign of confidence.