GM, EMEA @ Lorikeet | Tech laborer. Think too much about work | Behavioral Science too
Joined April 2015
- Tweets1.7K
- Following1.7K
- Followers676
- Likes4.5K
Robbie Tilleard retweeted
In the wee hours of this morning Sydney time we launched Lorikeet Dreamweaver. The reception has been awesome. Here's why we chose the name.
From the start, Jamie and I have been obsessed with how AI can unlock a completely new, different, and better type of customer experience.
Yes, being able to scale traditional inbound support is important. And helping companies do so has been our bread and butter. But we've always believed this technology is not a faster horse, but instead a vehicle for transformation.
Part of what gave me this conviction is that we *already* know how to create incredible experiences for customers - you see it all the time in enterprise B2B and luxury consumer. The problem was being able to provide them cost effectively at scale.
We called this product Lorikeet Dreamweaver after one of the best examples of this in reality: the restaurant Eleven Madison Park. In the period where EMP was the #1 restaurant in the world they employed people -- called Dreamweavers -- whose job was to deeply know their diners in order to craft bespoke experiences to delight them and create lasting memories [0].
We felt this described perfectly what we wanted to enable Lorikeet's customers to do for their customers. Giving every end customer an AI account manager enables this. The right message, the right help, the right advice at the right time. It delivers higher customer lifetime value, yes, but it does it by betting on the economic power of delighted customers.
What's more, by leveraging Coach (our AI Forward Deployed Engineer) we were able to make Dreamweaver a fully self-serve product that any business can try. We're so excited to see what everyone will do with it.
I find the 'writing is thinking' memes annoying
I think its making the same mistake as the engineers who insisted on writing every line of code by hand
Editing is thinking. Directing is thinking. Thinking is thinking. Don't be a snob about how to think
Robbie Tilleard retweeted
The 80 / 20 rules is still very much in effect. Accept the 80% has become so dazzling, no one wants to work on the 20.
Have been keen to see a similar sort of study on how models are influencing public policy.
If civil servants are all writing policy with AI - does it all revert to an AI-driven mean?
New Anthropic research: How AI agents shape experts' judgment calls
When experts delegate work to an agent, do they still make the key decisions? We tested this in a randomized trial with 50 professional data analysts.
1
118
Thinking about personalised learning and how good the models are now with motion and video.
Large opportunity to do specific videos for learning, add people as spectators into the narrative, create very engaging content. Layer over some good pedagogy and it’d be super cool
1
274
Just like marriages, no one can ever truly understand what is happening in other people’s careers
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined.
I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top.
The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever.
My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply.
To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had.
For me, it is time for a bit of a break before diving into something new.
If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
63
Best ways to get on the @dwarkesh_sp podcast:
1. Be born in New Zealand
2. Be born in Australia
3. Be Sarah Paine
66
The latest models are now very good at synthesis and analysis and turning them into good decks with simple prompts like 'no persuasion, civil service / McKinsey style'.
I'm allergic to slop but find myself editing less and less.
The factory came for coding first. Business analysis is next.
1
1
141
Robbie Tilleard retweeted
I think Ollie is right - probably the first time ever, globally, that an AI voice agent has handled a high $ value FX transaction fully end to end. Incredible start to our partnership with Eqauls
Robbie Tilleard retweeted
A friend who is now a Professor of Mathematics was always the best at maths in high school but he was also always the person who kept going until he solved each problem set. The bell would ring, we would pack up our bags for recess. He would stay until he'd finished the task and then he'd prepare himself for the next one.
That @Muse transcription and turn taking is very good. Here's why!
research.meta.ai/blog/introd…
85
I always liked @danhockenmaier's take on the requirement to use AI _and_ have strong judgement (e.g. read the thing / think about the thing)
Slop grenades works too from @tobi!
Lazy work used to mean too little output.
Now, with AI, it often means too much and more work for everyone else.
@tobi call it "slop grenades."
A "Slop Grenade" is when you let AI produce the work and pass it on without adding any value (including checking it).
Someone else has to wade through it, catch the mistakes, and clean up the mess.
You save time and look productive but someone else pays for it.
1
3
258
Robbie Tilleard retweeted
Jev is a textbook example of Clayton Christensen’s Innovator’s Dilemma playing out in AI.
It is commoditizing the bottom of the ML market: classifiers, routers, scoring and context triage. For frontier LLM labs, these workloads are not even worth pursuing. Revenue per decision is tiny, margins are razor-thin, and their entire infrastructure is optimized around selling expensive reasoning and token generation.
They cannot simply put a smaller LLM on fast inference chips like Cerebras and compete. Jev uses fundamentally different infrastructure, built to return typed probabilistic decisions so it’s a pure disruption
