I’ve been a copilot and ChatGPT subscriber for probably close to two years now, give or take a couple of months, and I had a trusted friend telling me for months to give Claude a try.
It took about two weeks of really running it through its paces, and constantly slamming against the limit on it to convince me I had to upgrade to at least the 100/month sub, and at this point I wouldn’t blink to bump that to the 200/month if necessary.
I 100% believe we’re in a bubble, and that this level of compute isn’t sustainable at this price point, but for as long as I have it, I’m going to run it at the redline.
I’m a solo dev working on a project that I’ve just gone full-time on, after about 1.5 years of part time work. It’s a codebase that I laid the groundwork in, and has very well established systems, standards, and constraints.
The work I’m using Claude to do is the exact work I would be doing myself, but it does it at somewhere in the neighborhood of 5-10x the pace I could have. I don’t know that I could get the same rate of production if I managed a team of 2-3 programmers. Right now, it’s literally almost perfect at taking my iterative suggestions, and implementing them at that accelerated pace.
Honestly the hardest part is dealing with the fact that at the end of the day, I have to understand this codebase perfectly (solo dev and all that), so I have to take in changes to it that are also 5-10x the rate my normal intuition would. But, again, the plus side is that it’s implementing them essentially exactly as I would have, as it has ~20k lines of code that I wrote to use as an example.
If I were to hire even one other programmer, I’d be paying well north of 5k/month, and I’d not only be managing a super computer programmer tool, but an actual human being as well. $100/month might as well be free comparatively.
Because I don’t believe 50% gross margins at face value, as being discussed in this thread I think the economics of all of these things are far more complex than that.
For what it’s worth, I haven’t staked my investment portfolio on there being a bubble, I’m just preparing for the worst and doing as much work as I can with Claude before a potential massive rug pull happens.
I'm not sure I've ever read something from someone so high up in a company that gave me such a strong feeling for "I'd like to work for these people". If job posts could be so informal and open ended, this post could serve as one in the form of a personality fit litmus test.
Once I gave this sometime I actually enjoyed the concept quite a bit. I think it's a novel twist on the bullet hell genre, and I'm a sucker for single button/input style games. I'm playing on a mouse, specifically a Logitech G502 that I can "unlock" the screen wheel to be almost frictionless, which to me feels like the right meta for a game like this.
Couple constructive criticisms:
- Is there a pause button that I'm missing? If not, add one, I want to actually play this more but I need to be able to take breaks
- The beginning is too easy. I'm biased as I like difficult games that encourage mastery, but I found myself dying early on runs because I was getting impatient with how slow the early progression is
- I haven't played enough yet to know whether monsters eventually come from the top, but eventually they should, it would greatly increase the difficulty curve
- Hitboxes feel a little meh, and don't seem to match the sprites
- If there's sound, it wasn't working for me, which leads to...
- Pump up the juice. You've got a great core loop, but you need the juice. One of my favorite game dev talks of all time would be perfect for you to watch and iterate with (https://www.youtube.com/watch?v=AJdEqssNZ-U)
- Monster diversity. Again, maybe I didn't play long enough yet to get to newer monsters, but from what I can tell the only diversity is their looks, health, and approach angle. Monsters that are faster and incentivize different movement, or that shoot back at you, or blow up after shooting so you have to avoid, etc.
I'll stop there, don't want an absolute laundry list. Keep at it, you've got a unique foundation, it'd be a shame not to polish it up!
Maybe add a charge meter that can fire a pulse weapon with a press of the spacebar after so many seconds of play. Unlock it later on in the game or something, idk.
I played for about 300 monsters and it got boring.
> Frustrated by the constant intrusions of child care and housework — “the back half of life,” as she called it — on her creative ambitions, she resolved to simply turn the work into her creative product. “Clean your desk, wash the dishes, clean the floor, wash your clothes, wash your toes, change the baby’s diaper, finish the report, correct the typos, mend the fence,” she wrote in the treatise, which was published by Artforum in 1971. “Clear the table, call him again, flush the toilet, stay young.”
There's more here than just the sanitation narrative, she's tapping into the focus Zen puts on the present, that these every day chores ARE the meditation, or in this case, the artistic expression.
I'm in the depths of optimization on a game right now, and it's interesting how the gains I'm making currently all seem to be a matter of scaling the concept of lookup tables, and using the right tool for the job.
What I mean is that traditionally I think peoples' ideas of lookup tables are things like statically baked arrays setup at compile time, or even first thing at runtime, and they never change. But if you loosen your adherence to that last idea a bit, where a lookup table can change slightly over time, you can get a ton of mileage out of a comparatively small amount of memory compared to wasting cycles every frame.
As for the right tool for the job, I've read tons of dev logs and research papers over the years about moving work to the GPU, but this last few months stint of ripping my game inside out has really made me see the light. It's not just lookup tables built at compile or early runtime, but lookup tables modified slightly over time, and sent to the GPU as textures and used there.
Follow this train of thought long enough, and now we're just calling memory writes and reads "lookup tables" when they aren't really that anymore, but whos to say where the barrier really lies?
To some extent it is required that all serious work by a computer be the kind of repititious thing that can be at least partially addressed by a lookup table.
If you take the example of a game, drawing sprites say. Drawing a single preloaded sprite of reasonable size is always cheap, so a slow frame must have an excessive number. It's very hard to construct a practical scenario of a large number of truly distinct sprites though. A level has a finite tile palette, a finite cast of characters, abilities, etc. It's hard to logistically get them all into a scene together, and even then it won't be that many. So the only scenario left where sprite drawing will be slow is drawing the same handful of sprites over and over again. By contrast that's super common: just spam a persistent projectile, tap the analog stick to generate dust particles, etc.
Without getting too much into detail (because the people I worked for were really paranoid, and I don't want to give them agita), we used to build lookup tables "on the fly," sometimes, deep inside iterators.
For example, each block of pixels might have some calculated characteristics that were accessed by a hash into a LUT, but the characteristics would change, as we went through the image.
We'd do a "triage" run, where we'd build the LUT, then a "detailed" run, where we'd apply the LUT to the pixels.
This mimics my experience. I bought the absolute bottom barrel M1 when they launched to replace my 2014 MBP, 8gb RAM and 128gb of space. The HD space is annoying, but otherwise this machine is untouchable. I do game dev work bouncing between the MBA and my gaming rig, which is Ryzen 7 2700, 64gb RAM and a 3070, and with certain benchmarks, the MBA still wins, silently, on battery for hours. Still blows my mind.
Is there a reasonable pivot for someone well versed in the software engineering space to get in, or is it still the playground of relevant Ph.Ds and the like? I've been up and down the stack from firmware to the cloud, going on 14 years in the industry, have a Master's in CS, am the technical lead for a team, yada yada, but have been flirting with the idea of getting out of standard product development and back into the nitty gritty of the space I first pursued during undergrad.
> Is there a reasonable pivot for someone well versed in the software engineering space to get in, or is it still the playground of relevant Ph.Ds and the like?
there's no such thing as a practical QC and there won't be for decades. this isn't a couple of years away - this is "maybe, possibly, pretty please, if we get lucky" 25-50 years away. find the above comment that alludes to "2019 estimates needing ~20 million physical qubits" and consider that this thing has 105 physical qubits. then skim the posted article and find this number
> the key quantum computational resource — are now approaching 100 µs (microseconds)
that's how long those 105 physical qubits stay coherent for. now ponder your career pivot.
source: i dabbled during my PhD - took a couple of classes from Fred Chong, wrote a paper - it's all hype.
The comments here are pretty funny, and do almost nothing for me more so than highlight the subjectivity of musical taste. I'd never claim to be an authority on music, but as the child of a professional musician and with chops of my own that I'd describe as "good enough to entertain myself and others at times", I love so many of the songs people are ragging on here, "We Built This City" included. Sure, there's plenty of stuff that I don't go out of my way to listen to, and again, subjective, but man, fluffy pop? Corny Christmas Music? Lewd Comedies/Parodies? It's all got it's time and place.
Tearing into a popular piece of media is just really, really cathartic, and much more so if a group of people agrees with you. And if we're gonna be honest its punching up most of the time, so it seems fairly harmless. Someone brought up Nickelback, its not like they've been shoved aside from relevancy, they were super popular the entire time people tore up their music critically. If someone puts them on at karaoke (or if someone puts on Starship for that matter) I'm sure most of the place is going to sing along.
For the climber, I thought the final solution was going to end up being rails that just pushed a weight further out from center mass at the top of the tower. The ball pendulum was cool nonetheless!
It took about two weeks of really running it through its paces, and constantly slamming against the limit on it to convince me I had to upgrade to at least the 100/month sub, and at this point I wouldn’t blink to bump that to the 200/month if necessary.
I 100% believe we’re in a bubble, and that this level of compute isn’t sustainable at this price point, but for as long as I have it, I’m going to run it at the redline.
I’m a solo dev working on a project that I’ve just gone full-time on, after about 1.5 years of part time work. It’s a codebase that I laid the groundwork in, and has very well established systems, standards, and constraints.
The work I’m using Claude to do is the exact work I would be doing myself, but it does it at somewhere in the neighborhood of 5-10x the pace I could have. I don’t know that I could get the same rate of production if I managed a team of 2-3 programmers. Right now, it’s literally almost perfect at taking my iterative suggestions, and implementing them at that accelerated pace.
Honestly the hardest part is dealing with the fact that at the end of the day, I have to understand this codebase perfectly (solo dev and all that), so I have to take in changes to it that are also 5-10x the rate my normal intuition would. But, again, the plus side is that it’s implementing them essentially exactly as I would have, as it has ~20k lines of code that I wrote to use as an example.
If I were to hire even one other programmer, I’d be paying well north of 5k/month, and I’d not only be managing a super computer programmer tool, but an actual human being as well. $100/month might as well be free comparatively.