Devan Vander

Fractional CAIO. Helping teams navigate the shift.
Ex-SFO | Ex-SoCal | Ex-Grand Rapids | Views my own.

[Devan Vander @1Principles_dev]

Sneak peak of my fireside interview at #LATechWeek

#SignalHelm #FractionalCAIO

[Devan Vander @1Principles_dev]

It's Conference Season!
Where everybody's talking about agents.
Nobody's talking about agentic.

#tech #ai

[Devan Vander @1Principles_dev]

I think it's important to have an non-government organization determine how we weigh what is moral and virtuous. Values are always weighted in our own equations, which is why we should aggregate those equations to do the most good.

[The Economist @TheEconomist]

Effective altruism is admirable, but dangerous. As the movement has grown stronger, it has become stranger https://t.co/1iP3N8nsYY

[Devan Vander @1Principles_dev]

They’re not shipping despite the pause. They’re shipping into the pause. The pause creates the attention surface that lets the new model land with the right narrative weight.

It’s deterministic storytelling in a probabilistic environment.

[Claude @claudeai]

Introducing Claude Haiku 5.5: the cheapest, fastest, and most capable small model we’ve ever released.

On average, it costs around 75% less to run than Claude Haiku 4.5.

[Devan Vander @1Principles_dev]

A lot of the frontier labs just said we need to slow down. Then they shipped new models the same week. How do I square that?
Essentially, the slowdown is the release: You’re giving the market time to absorb the previous generation while you surface the next one.
It’s probabilistic governance.

[Dario Amodei @DarioAmodei]

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.

Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.

You can read the full post here: https://t.co/OGyPb7yaYt

[Devan Vander @1Principles_dev]

Three years ago I wrote a post about why pilots don't reach production.

It got 40,000 impressions.

I still get DMs about it.

[Devan Vander @1Principles_dev]

@elonmusk true story.

[Devan Vander @1Principles_dev]

My DMs are blowing up after my review of #SignalHelm so I thought I would answer your question here:

What is SignalHelm?

[Devan Vander @1Principles_dev]

Three weeks with #SignalHelm. My honest review.

[Devan Vander @1Principles_dev]

@Joi2James ha! You just want to steal my ideas.

[Devan Vander @1Principles_dev]

5 things nobody tells you about AI adoption:

1. It's not about the technology
2. It's about the people
3. But it is also about the technology
4. Start small
5. Then scale

Save this. You'll need it.

[Devan Vander @1Principles_dev]

The 3 stages of AI maturity:

→ Probabilistic → Deterministic → Probabilistic again, but on purpose

Most orgs are stuck between 1 and 2. I help teams get to 3.

DM me "HELM" and I'll send the framework.

[Devan Vander @1Principles_dev]

A junior analyst asked me a question last week.

She said: "What if the data is wrong?"

I didn't answer right away.
I sat with it.

Because here's the thing.

The data is always wrong.

And that's not a bug.
That's the job.

She's 23. I've been doing this 14 years.
She taught me more in that moment than I've taught her all year.
↓
What's a moment a junior taught you something? 👇

[Devan Vander @1Principles_dev]

Correlation isn't causation. But at scale, it's close enough.