Victor Mota

co-founder, @coplane - the intelligence layer for the enterprise back office. previously: @Stripe, Google (BigQuery and Kaggle)

[Victor Mota @vimota]

If you're excited about helping define what the UX and pattern language for a future of agentic and self-assembling enterprise software, please reach out!

[CoPlane @coplane]

We're looking for NYC-based design engineers to join our small, talent-dense team.

This is an opportunity to define the frontier of HCI. We're building the platform for self-assembling, self-optimizing software already in production with consequential enterprises.

Our eng team is made up of folks who have built and maintained large open-source projects like Neovim and Analytics.js and scaled massive data systems at companies like Google, Stripe, Wix, and Segment.

DM if you are interested or know someone who might be!

[Victor Mota @vimota]

you can whine about ensloppification or you can take advantage of the generationally low bar that exists for everything now

when we post a job opening 9/10 applications are total slop, all it takes is one thoughtful email to get properly looked at - but no one puts in the effort

[Victor Mota @vimota]

there's a good chance society does away with factory farming and animal cruelty as AI gives us more empathy for non-human intelligence

[Victor Mota @vimota]

@zebulgar 100% this .. something about opus specifically gets people hooked

[François Chollet @fchollet]

Another thing: what you get from writing things yourself isn't just the code. It's an improved understanding of what the code does. That mental model is what lets you come up with further improvements, or invent a different way of doing things. You can't come up with ways to improve a blackbox you don't understand.

For most projects this doesn't really matter, because the code is the only thing you need. But if you're doing something novel, if you're doing research, the code is not the most important part. Understanding what the code does is the most important part.

[Victor Mota @vimota]

@patrickc But sad indeed! Used to love that place

[Victor Mota @vimota]

https://t.co/RWp8WBeJnk

[Victor Mota @vimota]

Firebolt just published a fascinating breakdown they completely rebuilt their data warehouse platform architecture with a verifiable agent harness for an agent-centric future.

They propose that for the agentic future - one where agents both develop and use the platform - database infrastructure should be monolithic and single tenant.

Interestingly, they also rebuilt their catalog and metadata service using the new Ducklake spec in order to benefit from an open standard and a verified oracle.

[dex @dexhorthy]

using AI for coding is a deeply technical engineering craft

most people don't approach it as so, and don't get the results we associate with high craft

but the ones who do have been sprinting ahead

more tokens wont save you, more thinking + skill + llm intuition will

have been saying this for almost 9 months now

[Victor Mota @vimota]

@thsottiaux True for codex and ChatGPT but I wish you could switch accounts on desktop and mobile (I have a personal and company account)

[Victor Mota @vimota]

When they announced that “feature” I was pretty sure they had misspoken.

Turns out it *actually* turns on the tokenmaxxing mode if you just say the word “workflow”??

[Matt Pocock @mattpocockuk]

So every time I say the word 'workflow' in Claude Code...

(let's say, when I'm creating a new GitHub workflow)

...it tries to enter 'workflow' mode, spinning up dozens of subagents to complete my task.

Stupid fucking thing

[Victor Mota @vimota]

@zoink 100% Ant’s models have way too much of a persona rather than being a neutral tool

https://t.co/uR689iu2xT

[Victor Mota @vimota]

Besides being much smarter, gpt 5.5 is a much more “neutral” model than opus 4.6.

You can always tell when a code change was written by opus. It tends to be extremely reward hacky and comment-ridden, whereas with codex, you can't actually distinguish it.

It’s much more steerable and lets the developers preferences shine through.

[Victor Mota @vimota]

@garybasin Good chance it’d be net negative in a lot of cases 😅

https://t.co/6VA1kHD74b

[Victor Mota @vimota]

Skill used to be a natural speed limit to what you could do in software engineering - but that's no longer true.

A junior engineer could only write so many lines of code that compiled and worked, and their skill was the rate limiter. As they got better the scope of what they could do increased along with the speed which they could do it.

With AI, that's no longer true. The raw output a junior vs senior engineer can produce is roughly the same, yet the quality and risk blast radius is vastly different between the two.

On the one hand, that's great - you're no longer bottle necked by technicalities that may or may not be relevant to what you're trying to achieve (ie. the particular syntax of a language).

But on the other, you now don't have any backpressure. Someone with lesser skill can produce as much in volume as an expert (which Chamath seems to think makes the latter no longer relevant !!) obsfucating the risk and tech debt that used to be naturally limited.

The biggest problem of all is that on the surface the two look the same - which explains a lot of the exec AI psychosis we're seeing right now.

[Victor Mota @vimota]

Expectation: "I'm sorry, Dave. I'm afraid I can't do that."

Reality: "Good pushback — you're right to question it. I overstated it. Want me to.. "

[Victor Mota @vimota]

Besides being much smarter, gpt 5.5 is a much more “neutral” model than opus 4.6.

You can always tell when a code change was written by opus. It tends to be extremely reward hacky and comment-ridden, whereas with codex, you can't actually distinguish it.

It’s much more steerable and lets the developers preferences shine through.

[Victor Mota @vimota]

From @mitsuhiko 's latest post on AI and the Pi codebase.

It's becoming more and more obvious that coding agents tend to just work in very incremental, short term fashion to accomplish any given task - as a result of the way they've been RL'd.

We collectively need to name this so we can work to avoid it. It's a special type of slop that needs its own term.

[Victor Mota @vimota]

@GergelyOrosz @zooko https://t.co/QiPR0fziXY

[Victor Mota @vimota]

Good articulation of a problem with AI not being self-limiting enough.

It feels like we haven't yet given a name to the issue of LLMs being too incremental and reward-hacky rather than pushing back on ideas and suggesting rework of existing code.

It feels like the core thing that leads AI codebases into the pit of despair.

[Victor Mota @vimota]

Good articulation of a problem with AI not being self-limiting enough.

It feels like we haven't yet given a name to the issue of LLMs being too incremental and reward-hacky rather than pushing back on ideas and suggesting rework of existing code.

It feels like the core thing that leads AI codebases into the pit of despair.

[Victor Mota @vimota]

Skill used to be a natural speed limit to what you could do in software engineering - but that's no longer true.

A junior engineer could only write so many lines of code that compiled and worked, and their skill was the rate limiter. As they got better the scope of what they could do increased along with the speed which they could do it.

With AI, that's no longer true. The raw output a junior vs senior engineer can produce is roughly the same, yet the quality and risk blast radius is vastly different between the two.

On the one hand, that's great - you're no longer bottle necked by technicalities that may or may not be relevant to what you're trying to achieve (ie. the particular syntax of a language).

But on the other, you now don't have any backpressure. Someone with lesser skill can produce as much in volume as an expert (which Chamath seems to think makes the latter no longer relevant !!) obsfucating the risk and tech debt that used to be naturally limited.

The biggest problem of all is that on the surface the two look the same - which explains a lot of the exec AI psychosis we're seeing right now.

[Han @HanchungLee]

holy shit where’s the evals to validate this holy shit.

[Victor Mota @vimota]

tech folks *want* people to know they're using AI, it's a flex 😅

[Victor Mota @vimota]

tbh I wish software engineers had a similar aversion to AI code. it's great to use AI, but I shouldn't be able to tell you used it.

my fiancee saying she thinks non-tech people are better users of AI because they don't want people to know they're using it

[rahul @0interestrates]

why do people (including me) have an aversion to AI writing but not as much to AI code? if a piece of text smells AI i stop reading it but i use things coded entirely with AI every day