Hacker Newsnew | past | comments | ask | show | jobs | submit | steve_adams_86's commentslogin

> If you leave your kid with a friendly neighbor

That's perfectly legal where I live, but... There are no friendly neighbours with that kind of time on their hands. My entire townhome complex of 34 units (ranging from 2–4 bedrooms each) has only one single elderly woman who's retired. Everyone else is young people in their 20s (working), or families with their own kids and—in all families but one—two working parents.


In my kids' lives, I've spent about $160k on childcare. So crazy to think of how hard we've worked to not be with our kids. I'm going to guess my parents spent under $10k in all of my childhood, inflation adjusted.

This model doesn't really make good sense. On one hand, I'm glad my wife and I can have careers. On the other, I doubt I would care much if we lived in a society where we didn't need to so badly. $160k of childcare doesn't pay for itself.


My understanding is that it's likely somewhere around 7.5% for men and women. Including bisexual people brings it closer to 10%. That's based on self-reporting, I think. I'm not sure how significant that would be in Meta's world.

Among men this would only be 3 or 4%. Probably not that significant given how coarse the strategy itself is.


Ugh, the gears and chain don't mesh and there's no sprocket on the rear hub

But seriously, I can't believe LLMs are able to one-shot a pelican on a bicycle this well. I wouldn't have guessed this was going to emerge as a capability from LLMs 6 years ago. I see why it does now, but... It still amazes me that they're so good at some things.


Is this capability “emergent”, or do AI firms specifically target SVG generation in order to improve it? How would we be able to tell?

I asked myself the same thing as I typed that comment, and I'm not sure what the answer is. I don't think models are specifically trained on this (though of course they're trained on how to generate SVGs in general), but I'm prepared to be wrong.

I have a feeling the most 'emergent' aspect was that LLMs have generally been able to produce coherent SVG for quite a while, likely without specific training at first. Since then I suspect there has been more tailored training because improvements have been so dramatic. Of course it makes sense that text-based images using very distinct structure and properties could be manipulated reasonably well by a text-based language model, but it's still fascinating to me just how well it can work.

Perhaps what's most incredible about it is how versatile human language is, even when it lacks so many dimensions as bits on a machine. Yet it's still cool that we can resurrect those bits at rest and transmogrify them back into coherent projections of photons from a screen.

I don't think LLMs are AGI or about to completely flip the world upside down or whatever, but it seems undeniably magical when you break it down.


Google specifically boast about their SVG performance in the announcement post: https://blog.google/innovation-and-ai/models-and-research/ge...

You can try any combination of animal on vehicle to confirm that they likely didn't target pelicans directly though.


next time you host a party, have people try to draw a bicycle on your whiteboard (you have a whiteboard in your house right? you should, anyway...)

human adults are generally quite bad at drawing them, unless they spend a lot of time actually thinking about bicycles as objects



Fantastic post, thanks for that.

What’s your point? Yes, humans fail sometimes, as do AI models. Are you trying to imply that, in light of this, AI is now as capable as human beings? If so, that conclusion doesn’t follow logically.

it's not a loaded point, i just think it's funny that humans typically cannot one-shot this. and it will make your friends laugh

And the left leg is straight while the right leg is bent.

EDIT: And the chain should pass behind the seat stay.


Agreed, multibuffers are such a huge QOL feature. I love being able to work across a dozen or more buffers at once with no impact on performance. You can work in so many places at once, navigate from the buffer to its file and back, widen the buffer up or down, etc. It feels like a super power.

What is the debt? As a user, Zed feels more like the only IDE that isn't weighed down with debt. It's incredibly fast, responsive, stable, and it's iterated on very quickly.

We've reached PGI

> no amount of static typing will save you from poorly defined or optimised-too-early types that encode business logic constraints into programmatic types.

That's not a fault of type systems, though.

> because business logic will move faster than whatever code you can write and fix, and exposing it as just "types" breaks the process for future programmers to extend your program

That's a problem with overly-tight coupling, poor design, and poor planning, not type systems

> In practice, I find that staunch static typing proponents are often middle or junior engineeers

I find people become enthusiastic about it around intermediate stages in their career, and they sometimes embrace it in ways that can be a bit rigid and over-zealous, but again it isn't a problem with type systems


> overly-tight coupling, poor design, and poor planning

yeah imagine if you could foresee the future five years in advance. People overestimate their ability to use type systems correctly, as shown in your reply here.


It has worked out well for me for 15 years or so.

It's so bizarre to watch. The programmers of tomorrow may actually use iPads and LLMs, or similar. The abstraction from the work being done is increasing rapidly, and understanding of the underlying systems is decreasing. I know it's the way things have been going for decades, but it feels weird.


I've always wanted to build one of these! This looks like great work, and I like the team-oriented premise.


Thanks Steve! The team angle came from mine and my co-founder's own frustration. We kept copy-pasting query snippets and results of queries into Slack/Teams to our engineering team but also Sales/Marketing etc. Figured there had to be a better way. If you ever want to give it a spin, I'd love to hear your thoughts.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: