Ralph Krauss

Software engineer / Entrepreneur

[Ralph Krauss @ralphkrauss]

@itsolelehmann There is even a much bigger gap between using AI in basic ways (as a chatbot) and using it in more advanced ways (agents).
It's a huge competitive advantage right now and I think it will remain so for the coming years if you stay up to date and use latest techniques

[Ralph Krauss @ralphkrauss]

@HealthRanger No need to overexaggerate… Claude models are still superior and their distribution is highly competitive.

[Ralph Krauss @ralphkrauss]

Microsoft authentication is the worst.
I have a problem with this every time I use the products.

[Sen @senbuilds]

Trillion dollar company btw

[Ralph Krauss @ralphkrauss]

@claudeai 👏Very happy to see this.
Even though gpt-5.6-sol might technically be smarter, Fable still brings something unique to the table.
It just seems to be much better at understanding what you are actually trying to do.

[Gustavo Noronha @gustavokov]

Linus jogou a real: AI é útil e vai ser usada no desenvolvimento do kernel Linux, quem achar ruim pode fazer fork e vazar

[Ralph Krauss @ralphkrauss]

@bslagter Dit soort angst zaaierij help niet in een tijd waar Europa zijn datacenter capaciteit enorm moet uitbreiden om competitief te blijven

[Ralph Krauss @ralphkrauss]

"we’ve essentially found a way to make sand think" it's so crazy if you think about that

[Demis Hassabis @demishassabis]

https://t.co/PTeDiv1b6L

[Ralph Krauss @ralphkrauss]

@kevinkern I saw a couple of posts already saying that using Terra is suboptimal. Not sure if that's true, but you could try using luna max as workers

[Ralph Krauss @ralphkrauss]

@melvynx What about terra ultra?

[Ralph Krauss @ralphkrauss]

New gpt model family explained

[Ralph Krauss @ralphkrauss]

People on X tweeting about how great gpt 5.6 is

[Ralph Krauss @ralphkrauss]

@theo Is it better at understanding WHAT to build instead of just knowing HOW to build it and going off?
And does it know when to stop and ask questions?

I think these were/are the biggest strengths of claude vs gpt

[Ralph Krauss @ralphkrauss]

@claudeai can we get a fable limit reset? obviously everyone was targetting the 100% fable use by today

[Ralph Krauss @ralphkrauss]

classic fable plays

[Taelin @VictorTaelin]

3am Fable

[Ralph Krauss @ralphkrauss]

@forgebitz The hard part is designing the ai workflows, creating the required pieces to give your agents smooth sailing.
The easy part is now becoming an 'operator' and overseeing the agents do the work and correcting where needed.

[Ralph Krauss @ralphkrauss]

This is the first thing I'm setting up today. Having Fable as the orchestrator with GPT-5.5 as the workers is really the perfect combination.
Claude is just better at understanding your intent, and asking the right questions. And GPT at executing it.
And hopefully this way we will not immediately hit the Fable token limits.

[Claude @claudeai]

Introducing Claude Science, a new app designed with every stage of research in mind.

Artifacts traced to their code, environments managed on demand, and 60+ optional scientific databases that you can connect.

Available now in beta.

[Ralph Krauss @ralphkrauss]

Yes. I do this in slack already with hermes and it’s much more customizable to my workflows and very easy to swap models.
But don’t forget that people value convenience. Anthropic will probably create this nice ecosystem where knowledge gets shared across all their tools which will make it super user friendly.
Similar to what Apple is doing

[Ren @ren_stocks]

Satya Nadella just posted something that validates the entire AI buildout thesis from the very top of the stack.

The model is commoditizing. The durable value is the learning loop a company builds on top of the model.

He splits it into two assets:

Human capital -- the knowledge, judgment, relationships, and pattern recognition of your people.

Token capital -- the AI capability the firm builds and owns.

He says the real opportunity is building a learning loop where human capital and token capital compound together.

If the model layer is commoditizing then the durable returns are not in the model makers. They are in the infrastructure that powers every company building its own loop. Compute. Memory. Interconnect. Power.

The full stack underneath the application layer.

The model wars will have winners and losers. The infrastructure underneath gets bought either way.

Bullish the AI buildout.

Every layer. If you want to understand them in detail, check out my Substack.

https://t.co/Wna5UzCOVT