Hacker Newsnew | past | comments | ask | show | jobs | submit | hamish-b's commentslogin

Excellent course (I did mine ~2019), super grateful to Paul and Stanford for continuously releasing it publicly and updating them all these years later.


What a wild ride, the traffic to my site is more akin to a rollercoaster. Got better for a few mins and then fell back apart.


I'm surprised people aren't really talking about how this affects Adobe going into the future. Adobe frankly has been hostile to it's users for the past while, this should really shake up the game. Creatives now have access to a pro-level tool for free with the option to pay extra for AI features. This is a clear shot across Adobe's bow and positions Canva to control the creative vertical from professionals to the average person.


Site is currently down, wayback machine link: https://web.archive.org/web/20250919031347/https://lenowo.or...


yeah, hackernews killed my poor ip camera again xD


You should get more cameras and run one of them as load balancer in front of the others!


I actually wanna make a cluster at some point


a beowulf cluster? remember those? :-)


I want to make a CPU generation beowulf cluster one day. I'm thinking a Pentium MMX box, a K6-3 box, a Pentium 2 or 3, maybe even a 486 if supported. Then have some compute job that runs across all of them. :D


Nice work, what you're up to is pretty impressive :^)


My problem _still_ with all of the codex/gpt based offerings is that they think for way too long. After using Claude 4 models through cursor max/ampcode I feel much more effective given it's speed. Ironically, Claude Code feels just as slow as codex/gpt (even with my company patching through AWS bedrock). Only makes me feel more that the consumer modes have perverse incentives.


I almost never have to reprompt GPT-5-high (now gpt-5-codex-high) where I would be reprompting claude code all the time. It feels like its faster, doing more, but its taking more of the developers time by getting things wrong.


It’s great for multitasking. I’ve cloned one of the repos I work on into a new folder and use Codex CLI in there. I feed it bug reports that users have submitted, while I work on bigger tasks.


I've been working hard at this problem over at https://kurnell.ai. Our thesis is that going forward it's unrealistic to ban or detect-and-reprimand AI, therefore we need to meet students where they're at and democratize access to the best AI for all.

We have great traction with universities in USA and Australia. The flywheel that we've constructed means that students are being prepared for industry + research in a Post-AI world, and professors can see exactly how students are using AI tooling. Our findings are that knowledge of how students are using AI goes a long way to helping institutions adapt.

Keen to chat and share our findings - reach out at hamish(at)kurnell.ai !


Note that the article references 65B -- this is in AUD.


Extremely well said, as someone from the inner west suburbs I totally get it. Brisbane is having it's time in the sun, we deserve it!


I had hand-done mine on https://kurnell.ai - its really cool to see that I could have computationally done this instead :)

I used rive.app to encode the frames and create a state machine to move between the states. Perhaps I can simplify this even more.


Super cool! I don't see any clear description of what the difference between the pro and the free licenses are. Maybe the website could be updated with that?


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

Search: