jenniferhooley

I loaded my smallest professional project today with fairly simple shapes and text and a minimal number of large smart objects with few effects (essentially 1 page of an educational web system design).

It's not as bad as a few days ago, however it's extremely slow and lots of little bugs to the point I would never use it professionally. Weird things too, like my entire mouse pointer is slow, drawing with a brush is slow and jaggy, editing a vector and the points don't highlight properly or ending the edit and suddenly it randomly moved somewhere on the screen. Just very clunky, bad feeling and not really usable compared to photoshop.

I'll keep an eye on it out of curiosity, but right now it's unusable in a professional context.

Yeah agreed, I was talking to a friend few years ago when Google first wiped out blogging by taking over the blog content and showing snippets instead of real links.

Obviously the end-game is for the companies to automate all the text, games, and software so you just go to their feed and pay them a monthly instead of individuals making making this content.

No, this is exactly what Claude vibe codes, this is definitely a one-shot Claude style game (assuming it looks up a decent pixel art pack). LLMs are pretty decent these days at making standard game clones like this, you just have to add a system prompt about making it look like a real indie-style pixel art game instead of just letting it make it's standard "web game" style. I assume google has a good background system prompt/harness to do this automatically.

Have you tried the apps, or have any experience with Photoshop (complex photo or art or video projects)?

Just curious, because I've worked within design/film etc. for years.

I did open the PhotoCraft and loaded a few of my simpler layout projects. And literally nothing works.

Let's see:

1. Tried to select text, like the selector is off from the cursor.

2. Tried to edit text, takes like 3 seconds to drag the slider for size, nothing happens (text doesn't change size).

3. Absolutely everything is insanely laggy: selecting, switching, etc.

4. Can't seem to expand/collapse folders.

5. Everything shows shortcut keys - shortcut keys don't do anything.

6. Tried to move something, it kind of got frozen for a while. Eventually disappeared I think?

7. Clicked something and it auto selected a side panel (don't know why), hid the other panel, can't stretch, collapse the panels or get it back...

This is actual joke software, I'm confused why you are impressed?

Without going through the following flow I very quickly end up with nitty recurring micro bugs or interaction "gunk" and 10's of thousands of lines of pointless code that bloats out the code-base and agent's context.

plan -> approve -> agent codes -> review code -> cleanup and pointing to specific skills/agents depending on what the issues are and some manual instructions pointing to specific lines of code-> agent codes -> review code -> final cleanup -> ship.

Working on a 3d sims-like video game I cannot get away with much less than that flow outside of very small features. I have (like all of us I assume) tried moving forward over weeks of not code reviewing and only plan reviewing, and it's amazing how badly things fell apart, I ended up having to reset weeks of work and ended up knocking out like 50K loc for the exact same features and zero bugs instead of constant bug (or just interaction/latency gunk cropping up everywhere) cleanup slowly escalating and agents needing to take hours to build anything.

Might find this interesting: https://earendil.com/posts/measuring-code-sloppiness/

Maybe not... I personally wouldn't use a service that would boot me off if they did find out because I spoke about it online or committed too much code one week or something.

As much as I hate vibe-coded "slop". I feel like this is idiotic. I and most good engs with decades of experience do use AI now to type nearly all the actual "code", but we read EVERY SINGLE line, always fix and rework every feature before a commit to make it clean, optimized, and de-slopified.

But still, we wouldn't be allowed to use codeberg because "generative ai" was used significantly for most lines of code?

lmao. Yep. I guess they all took a sabbatical and decided F** it, lobotomize me, I'm all in.

No they don't know the domain at all. I think one of the startups brings in around $10K a month now. They know nothing about the domain only that it brings in money (i.e. they run thousands of startup "tests" and this one seemed to be easiest to make money from), and even less about code (have no idea what a variable is).

I just think devs are coping by thinking that saas startups have any barrier to entry at all. You are now competing with every single person on the planet who wants to make money sitting at home on a computer, and likely the business guy with zero skill or craftsmanship is going to beat you. The only thing with barrier to entry is insanely complex software (e.g., what used to take a genius team of 10 multiple years to make) and or more creative software like video games (where you need extreme care and taste to get something anyone will look at for > 2 hours).

Average saas products require no skill or craftsmanship and literally anyone including our grandmas can make it (assuming they will take a few months at least to chat with an LLM). To be clear I think this is great! More better software, people with smaller needs can have llms build it for them totally bespoke. For those of us with skills we have to move WAY out of our old comfort zone and make way cooler things, a standard saas that takes a few months isn't something we should waste our time on anymore.

Not really true tbh. Plenty of 50 year old business guys with less than zero technical ability making saas startups over months of focused chats with LLMs. Of course they're slower at making stuff and the projects are worse than a smart engineer, but they can definitely ship and run to thousands of real customers.

Do you really think there aren't data annotator services specifically training for SVG drawing? And that those annotators don't have a rich set of frequently requested icons/graphics/etc. that they review and train on?

I'd be shocked if they didn't myself.

Do you think young children should never be in a car either? There is always a chance of death, and at some point you have to let your kids enjoy life, some of my best memories are at that pond watching dragonflies and listening to the water splash. It just seems so sad to not allow a child to be free to enjoy the natural world, yes a very scary speech about the dangers is important, but at some point it's important to let children learn on their own I think... I don't know... I'm just really glad I had the freedom and always remember my childhood alone in the woods and at the farm with deep happiness that I never felt locked up in a house or classroom despite those being super safe.

GLM 5.3 (usually flash) + OpenCode.

Myself and my siblings all lived on a fairly large rural farm with 2 large bodies of water that we always hung around unsupervised from about 4 y/o onwards. We loved catching bullfrogs and watching the fishes and stuff. Our mom gave us very strict/scary warnings about never going in and being careful, but other than that we were pretty free thankfully. 5 y/o with a good warning seems totally fine.

I'm using chrome on windows (using mouse scroll wheel), that page is super slow with an annoying easing function.

It definitely seems to be.

However, if it helps to move Linux into the mainstream desktop choice. I'm all for it. Obviously I won't be using it any time soon - but if some rich dudes want to throw millions into marketing and getting Linux mainstream, seems like a net positive to me.

Huh? I have a 650sqft house, it costs 430K in a small town. I looked at building a new place, it would cost roughly the same to hire a building team to make it - not to mention years of waiting on development. The land is fairly affordable if I buy a little bit out of town. It's the cost of building that is stopping me.

Or we just wait for automation and robotics to make building housing much cheaper and the free market system to fix things - which 5 years ago I assumed was too far out, but with all the advances in the sector, seems like we might get significantly cheaper building costs within the next 10 years or so.

Here I will quote what I said to you - it's super explicit and obvious I'm asking how one would find it useful.

"How is it useful? I'm no expert in combustion engines, so I'd hesitate to rely on a slop graphic like this to learn about them as I doubt the creator has any idea about them either."

Notice I say "How is it useful?" and "I'd hesitate to rely on a slop graphic like this to learn".

I think I'm being very obvious about the fact that I can't understand how one would find it "useful" to "learn". Absolutely NOWHERE did I say you can't be entertained by it.

Where did I say what people are allowed to find entertaining? It would be insane to tell someone what they should find entertaining, that's the dumbest thing I've ever heard.

As for useful, I'm genuinely trying to understand HOW you find something useful that may or may not have zero accuracy to the thing in which it's portending to showcase?

When I don't know about something - I would find someone who's really knowledgeable (or a resource that's vetted), and learn from that obviously. I don't want to learn something completely unverified and made up? That makes no logical sense.

Just like my friend's slop graphics. I imagine learning geometry from that and fundamentally locking in incorrect knowledge when there are many thousands of resources that would teach me correctly - this makes no sense right? Would you prefer to learn from the pretty graphics that are teaching you nothing?

How is it useful? I'm no expert in combustion engines, so I'd hesitate to rely on a slop graphic like this to learn about them as I doubt the creator has any idea about them either.

It's like a friend - he was like "Look at this awesome set of categorized interactive animations from claude to learn geometry!".

And like 60% of the animations were technically just wrong. Very pretty though, and the effects were cute... I guess...?

Oh, you think we are going to have another breakthrough in artificial intelligence rivaling the one we had in 2022 this year or next? What research or studies lead you to believe this? Could I see one?

I haven't personally seen the research leading me to believe there's a massive new intelligence breakthrough on the horizon but maybe I missed it!

As to the current version - I use AI constantly, as in 12 hours a day 7 days a week specifically for high-end game development. It is currently so far from the 2027 prediction that it's hilarious if you think a little incremental improvement (e.g., like from opus 4.6 to Fable) is going to bring us there.

No. Anyone who doesn't code with AI - while retaining a deep knowledge and understanding of the problem domain - is falling behind.

I hate to say this tbh, I loved hand-writing code. I made a great living for 20 years, and I absolutely loved it and was quite good at it.

Hand-typing code is just slower now; there’s no two-ways about it. You are either going to be slow and a bad hire for businesses, or you figure out how to adopt AI into your workflow to speed up.

One thing I think people don't realize is that deep knowledge of programming, performance, architectural, and domain specific trade-offs makes a skilled engineer about 1000X faster than someone without those skills -with AI. But yes, now unskilled people can actually make apps/software. They just tend to be slow, and their products are full of bugs, security flaws, and abysmal performance.

So we went from: Skills = can or cannot ship any software at all. Now we are at: Skills = can ship better software much faster than unskilled people.

I was actually faced with this recently. I decided to learn Rust and port one of my side projects to it. Initially, I moved extremely slowly, and the AI made truly horrific architectural decisions because I didn't have the knowledge of how to direct it, especially compared to my primary languages.

However, once I gained a firm grasp of Rust, I was better able to properly direct the AI to fix fundamental issues and architect things properly. My speed increase multiplier proved to be directly proportional to my growing knowledge of both the language and the domain.

Skill and knowledge combined with AI, when used appropriately, absolutely multiply your speed and quality. I really think once you understand what AI can do, and how to utilize it to produce better code, faster than before, there truly is no going back.

I'm finding a path forward that I actually enjoy now and don't really see losing my value (no telling how things will change in the future), I can have more time to focus on really quality/solid/performant and useful systems with less time just typing one character out at a time.

You could have talked to me 3 months ago and I'd never imagine I'd say the above btw. I REALLY enjoyed code writing and earlier AI models without harnesses were pretty useless for anyone skilled at development. Now with stuff like deepseek Flash I feel like I have a happy medium of 100% directed/fast code turnaround, less typing, more deep focus on architecture, systems, and the actual end product.

Programmers can use smaller models like deepseek v4 flash for 98% of the same productivity as SOTA models and cost (true cost) around $10-$30 a month. So I doubt most people who heavily use them are too concerned. It's only vibe/hobby coders who really need SOTA and they probably don't think about it much.

Man the one-shot game that is genuinely good in 2027 is crazy. A good game typically takes around 3-8 years of development by multiple skilled people. Maybe 3 years or so for 1 guy that's super dedicated.

Right now single prompt with Fable can get us a small protype of like 1 game mechanic that's not even remotely production ready. So this guy thinks it'll 100,000X in a year.

AI boosters are something else.

Let's say we're prototyping an interactive tree (this is totally made up, but you get the idea):

Take this data input and convert it to a Sugiyama-style tree with hand-drawn feeling lines connecting the nodes. We need the ability to activate a random subset of nodes with a paint splash. The whole thing should feel organic, incorporating small motion effects. As the nodes are activated, the edges should look like a hand-drawn painting effect drawing toward the node, and then SPLASH onto the end node as it changes from black to deep red. The background should utilize a muted paper color, and we must adhere to this color palette for all elements (PALETTE).

...then back-and-forth 10 prompts or so to get the prototype I was looking for. Comparing these types of things between Fable and Opus, something like this would be quite a bit less glitchy, prettier, and closer to the quick prototyping I needed than what I got with Opus 4.8.

Now, when I went deep into a complex codebase to fix a small issue or optimization that spanned many files and was fairly unique from anything in the training data, I didn't really see any noticeable difference between Fable and Opus 4.8.

I felt like it was about 10X better at "pretty" but straightforward 1 shot'ish type tasks. Not so different for complex and specific tasks in real code-bases.

Why do you say it was a lot better, what type of tasks were you testing it on?

Yes, but the AI will handle asking you everything. Here's how easy it is to make a game that functions like his but without the polish - put Claude in plan mode 1 prompt, then answer any questions it has:

--- Make me a threejs game in html. The "game" is where you can tap a button to add a log to a cutting block. You have an axelike (e.g., a stem cube for handle and slim cube for axe blade) object that you can axe into the wood, and wherever the head intersects the underlying "log" cylinder it splits the log.

So blade is a hittest object, where that object hits the spawned log, the log splits, use a decent physics lib (ammoJS is fine) so it feels good. Do you have any questions? ---

After answering about 5 basic questions, the above made a cute little physics based log cutting game very similar to this guy's demo - now it'd probably take about 5-8 more hours to slowly prompt out a more polished version. This stuff is really easy - you don't have to have any experience in game dev to prompt it up.

You mean by "we illustrate and print it", we have an AI illustrate it?

How much of the code do you think is written by AI? Just curious as I do video game development and just recently started heavily using opencode agentic development with the Flash models (instead of essentially using zero AI assistance before). I actually really like the workflow and find it helps speed things up a hair and in general allow faster testing/tooling builds/etc. which are super helpful to make a game feel really good.

Kind of curious how other people are using agentic code tools for game dev!