feuter

Imagining a future @rubiconpay @mutualprog

[feuter @feuters]

@wenkafka @RubricLabs sick

[Diogo Almeida @CompleteSkeptic]

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?

I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev

• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions

AFAICT the shortest path to AI-based economic revolution

[Canteen @thecanteenapp]

🥈 2ND PLACE (pool of 2)
PULS @pulsmarket — @rdmnad
RUBICON @rubiconpay — @wenkafka & @feuters

🥉 3RD PLACE (pool of 3)
AGORAFX — @Musa_Ais
ATHENA — @Suganthan2006, @sylesh_7 & @Austrix_XD
ENGYE — @dadadave80

[feuter @feuters]

In Silo season 1, Juliette listens to the generator and knows a rotor blade is defective. Years around that machine had built a world model in her head.

While this is cool and all, you can't clone Juliette. Two problems stem from this: increased dependency and scaling issues.

Industrial systems are growing too fast to stay tied to tacit knowledge.

I believe enormous resources will go toward physical AI to close that gap. Models that learn from vibration, temperature, sound, and video. We have to turn a technician's intuition into a model that can actually scale.

[feuter @feuters]

Peak Montréal location (literally)

[Rubicon @rubiconpay]

We integrated with @deck_agent to make importing from Substack effortless.

Writers can let an AI agent migrate their content without lifting a finger.

Launching this week.

[feuter @feuters]

@rubiconpay Card fees just aren't built for micropayments. Good to see Franklin Templeton saying it out loud.

[feuter @feuters]

We live in the attention economy and all these labs are fighting for narrative position.

[Claude @claudeai]

Beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans, at 50% of limits.

Pro and Team Standard users will continue to have access to Fable via usage credits, and will receive a one-time $100 credit.

Demand for Fable has been challenging to predict, which is why we rolled it out to subscription plans in stages, extending access several times as we secured additional capacity.

[feuter @feuters]

@0xVatsaShah @Valiron_grade @SuperteamUSA Good stuff

[feuter @feuters]

We tend to think predictions are passive expectations waiting for an uncertain future to crystallize. But credible predictions become actors of transformation.

People build around them.

Moore’s Law became real partly because the semiconductor industry organized itself around it, treating it less like a forecast and more like a target.

[feuter @feuters]

Christmas morning

[Claude @claudeai]

We're extending access to Claude Fable 5 on all paid plans through July 12.

[feuter @feuters]

@andruyeung @Cloudflare Agents can pay creators for what they read. Close the gates, agents can be customers, not just scrapers.

[feuter @feuters]

https://t.co/t2IBLCveFY

[feuter @feuters]

@isaaccyn Love it

[feuter @feuters]

My thesis is that if you set out with the goal of finding a problem, you'll start seeing problems everywhere, whether or not they're actually there.

Think of confirmation bias. Once your brain is set on looking for pain points, it starts interpreting neutral or minor things as if they were significant ones.

[feuter @feuters]

https://t.co/t2IBLCveFY

[feuter @feuters]

@rubiconpay @artemis Rubicon is the stream. Discovery is the boat.

[feuter @feuters]

@wenkafka @artemis @rubiconpay With practice you might get 15

[feuter @feuters]

Any company can bolt on an LLM and say "Let's raise as an AI company". Robotics has hard barriers to entry, fewer teams can attempt it. So deal supply is structurally thinner, regardless of how hungry VCs are.

[Andrew Kang @Rewkang]

VC interest in robotics has barely started

[feuter @feuters]

@marketplunger1 Increasingly making genuine human writing more valuable

[feuter @feuters]

After a week of chatting with creators, we realized one of the biggest friction points was onboarding. Nobody wanted to spend 5 minutes setting up their article on Rubicon.

So we brought it down to 30 seconds.

[Rubicon @rubiconpay]

Make your Substack writing discoverable to agents in under 30 seconds, and get paid when they read.