Matt Cline

I make software

[Matt Cline @mattcline__]

@naval "slow is smooth, smooth is fast"

[Matt Cline @mattcline__]

@andruyeung leetcode

[rygo6 @_rygo6]

I still believe a correct framing on LLM usage is that they can get you 80% of the way upfront, but that last 20% starts to become impossible to overcome. The proper way to use an LLM is to do the difficult manual 20% upfront. Once you have some decent structure, decent API surface area, decent data structures laid out, decent base libraries. Then you can largely prompt your way through the last 80%. While the first 20% which you did manually, intentionally and fully comprehend. That 20% essentially becomes part of your development harness, almost a prompt itself. Where you can steer its overall architecture, or data design, or other critical pieces by keeping a grip on that first manually done 20%.

Something real concerning with this. Knowing what to do in that first 20% requires advanced knowledge of the problem domain you are going to deal with. It is a staff or principal level engineer domain. Which most simply can't do.

Which means most people are going to run the gauntlet heavily relying on prompts, and are going to cover some incredibly impressive ground very quickly. Because 80% of the way is quite a long ways there. It looks like it could almost be done. But then they will get stuck in that last 20% and will never be able to make it perfectly polished. Nor perfectly sable. Nor remove all the odd quirks. Nor overcome the last pieces of performance or memory issues. It will basically hit a plateau where they lose control of it. Or it becomes unwieldy to take it in a new complex direction, or add some novel complex functionality to it. It will be very hard to tell from the outside if any given company or project is aimed at getting stuck in that last 20%. As they could appear to be kicking some major butt in that first 80%. But then it will reach a plateau where they can no longer do anything substantial without breaking something else, or seem unable to fix every little quirk or issue. A lot of companies will probably get stuck here. With dozens of very impressive things 80% of the way, but seem unable to get over the last final 20% to really perfect it and polish it.

[Matt Cline @mattcline__]

@artie_labs Artie

[Artie @artie_labs]

We offered 4 engineers $100,000 to break Artie.

They tried everything.

Schema changes. 10M-row updates. TOAST columns. Database version upgrades. Data type changes.

Watch what happens 👀

Artie is a real-time data pipeline that streams your database changes into your warehouse with sub-minute latency and no infrastructure to manage – built to handle what actually breaks in prod.

Sign up today and try it yourself: https://t.co/fcCL14mJ2s

Repost and comment ‘Artie’ and we'll DM you $10,000 in credits ⬇️

[Matt Cline @mattcline__]

The accounts I follow on X act as high-quality filters of information.

[Matt Cline @mattcline__]

Don't reinvent the wheel - instead focus on building the next bike frame, the next application of the wheel

[Matt Cline @mattcline__]

There is no finish line

[Matt Cline @mattcline__]

@SpaceX @Starlink Where did the booster land this time?

[Matt Cline @mattcline__]

Can now chat with AI alongside the document editor. Check it out at https://t.co/mVoAwUrRRU

[Matt Cline @mattcline__]

I made a tutorial about building a collaborative and persistent text editor using Lexical, Yjs, and Postgres.

Full video: https://t.co/LJMM3GHBpI
Blog: https://t.co/W5nxARgxjf
Source code: https://t.co/LiX0GUDG5y

[Matt Cline @mattcline__]

All about the timing

[Tony Dunn @tony873004]

The ESCAPADE mission to Mars is about to launch. It will stay near Earth for about 1 year before using an Oberth maneuver to send it to Mars.

[Matt Cline @mattcline__]

You can now share and edit documents collaboratively! See it in action: https://t.co/mVoAwUrRRU

[Matt Cline @mattcline__]

@huggingface has a library of free courses to learn about LLMs, MCP, etc.: https://t.co/pUEaf2Ozzw

[Andrej Karpathy @karpathy]

How to become expert at thing:
1 iteratively take on concrete projects and accomplish them depth wise, learning “on demand” (ie don’t learn bottom up breadth wise)
2 teach/summarize everything you learn in your own words
3 only compare yourself to younger you, never to others

[Tony M @twowheelsoneman]

@valigo Peter Thiel said it best. Any subject called "science" is an embellishment, like CS or Political Science. Chemistry and Physics are self-evident practices.

A better description would be Algorithmonomy or "the order of understanding step-by-step procedures"

[Matt Cline @mattcline__]

I love this question. Before starting my own startup, I would choose 'envision and build' in an attempt to avoid the mess of deconstruction. Now, I choose 'deconstruct and evolve' since it implies starting from something that already works.

[Matt Cline @mattcline__]

Which creative type are you? I'm the alchemist - "You’re drawn to deep inquiry, stripping away illusion and revealing hidden truths." #mycreativetype
https://t.co/BXfxXQ1YtI

[Matt Cline @mattcline__]

@naval Ownership is the foundation of compounding.

[Matt Cline @mattcline__]

@naval Ownership is the foundation of compounding.