Technologist of 20 years. Having fun building with agents, exploring critical rationalism and being an amateur endurance athlete

Sydney, New South Wales
Joined January 2026
Include

Only show posts containing:

Exclude

Hide posts containing:

Time range
-
Minimum likes
Any news on when Quint Studio might land?
14
Replying to @0xjmori
@bot what can we learn from this?
147
Replying to @Hello_World
Except they don't think
1
1
2
84
Replying to @Hello_World
Is Bill Gates still relevant?
4
Love building with @spacexai Grok Bot. One feature request: let a single bot run more than one browser task at a time. Mine has become extremely good at queuing. @poteto @elonmusk any chance? 🙏
4
LLM's don't deal with ambiguity well the chain you described moves from more to less abstract. A sufficiently detailed spec is code
79
Replying to @elonmusk @bot
@bot tell my chief about this, he's not aware of it
1
1
42
You're the perfect antidote to @Steve_Yegge 😂
4
3,130
Replying to @unclebobmartin
How do you find the cursor cloud agents? For me they're like blackholes. Really want to transition to them but can't get anything done with them
2
2
1,691
Replying to @mattpocockuk
What's your view on the future of factories? Do we all have our own instantiation of one made of composable units much like DevOps today? Or do we coalesce around patterns and practices that works and buy factories off the shelf?
1
1
876
Replying to @thorstenball
I agree with almost all of this except for "Stop worrying about accumulating slop". If I provide the why the what ends up looking like 10's or 100's of .py files in the root of the project with very little cohesion. What are you seeing that's different?
301
Replying to @JackEllis
I found cursor cloud agents very slow unproductuve and expensive and i have a 300x plan. What am i missing?
5
Replying to @milesdeutscher @bot
@bot see if any of these uses cases are relevant to us
1
496
Replying to @poteto
Love this! I see pstack already maps model actions to real code, tests with deliberate breakage, and uses statecharts for human review. The gap I’m curious about is completeness: how do you catch a requirement the model missed, even when TLC passes? Any ideas beyond review?
89
What If We’re Building AGI Backwards? simon336766.substack.com/p/w…
2
3
22,848
Replying to @DavidDeutschOxf
@grok who's quote is this?
1
1
1,006
Replying to @poteto
@bot let's digest this post and the prior one in the series and see how we might incorporate some of its learnings into our factory
1
13
Replying to @0x_rody
@bot help me understand this post. How could I start using this?
600
@bot summarise the key takeaways from this podcast and help me decide if we should make any changes to our development approach
1
237