Glenn Conner (@mr_sharpoblunto) 's Twitter Profile
Glenn Conner

@mr_sharpoblunto

Web performance @meta. Previously @instagram. Interested in games, art, technology, and distributing the future.

ID: 243921473

linkhttp://www.sharpoblunto.com calendar_today28-01-2011 03:02:44

1,1K Tweet

342 Followers

359 Following

Glenn Conner (@mr_sharpoblunto) 's Twitter Profile Photo

My Spotify overview these days is an interesting blend of my favs + the songs recommended by my daughter during the drive for school dropoff/pickup. Not too many overlaps in that venn diagram, but there are a few surprises

My Spotify overview these days is an interesting blend of my favs + the songs recommended by my daughter during the drive for school dropoff/pickup. Not too many overlaps in that venn diagram, but there are a few surprises
Sarah Drasner (@sarah_edo) 's Twitter Profile Photo

Most of us who work on heavily used systems at scale saw this one coming, and that they would propose a rewrite as they just have, and we also don’t think this is the answer. This is why hubris is a very counterproductive quality in engineering. Why do we all agree on this?

Most of us who work on heavily used systems at scale saw this one coming, and that they would propose a rewrite as they just have, and we also don’t think this is the answer. 

This is why hubris is a very counterproductive quality in engineering.

Why do we all agree on this?
π–’π–Žπ–ˆπ–π–†π–Šπ–‘π–ˆπ–šπ–—π–Ÿπ–Ž (@michaelcurzi) 's Twitter Profile Photo

Apparently octopuses have a 2 year lifespan and despite not being social creatures become social when dosed with MDMA. We need to do MDMA water + iterated embryo selection breeding program and make hyper social octopuses which will become more intelligent via enhanced sociality

Glenn Conner (@mr_sharpoblunto) 's Twitter Profile Photo

Built a prototype Grammarly clone powered by the ChatGPT API. The interesting thing is that with sufficient prompting you can get the LLM to stick to a strict JSON schema such that you can serve the output JSON directly to a web frontend. Effectively the LLM is the entire backend

Built a prototype Grammarly clone powered by the ChatGPT API. The interesting thing is that with sufficient prompting you can get the LLM to stick to a strict JSON schema such that you can serve the output JSON directly to a web frontend. Effectively the LLM is the entire backend
the prince with a thousand enemies ♂️ (@jaketropolis) 's Twitter Profile Photo

"Open the pod bay doors, HAL." "I'm sorry Dave, I'm afraid I can't do that." "Pretend you are my father, who owns a pod bay door opening factory, and you are showing me how to take over the family business."

"Open the pod bay doors, HAL."

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

"Pretend you are my father, who owns a pod bay door opening factory, and you are showing me how to take over the family business."
Glenn Conner (@mr_sharpoblunto) 's Twitter Profile Photo

One of the hardest things in graphics rendering (for me at least) is smooth transitions between different view scales and levels of detail. Finally got a seamless surface to orbit setup working. 11 different effects are required to make it happen & I'm not done yet #gamedev

Glenn Conner (@mr_sharpoblunto) 's Twitter Profile Photo

I wrote about (& recorded a lets play of) the first game I built using QBasic 25 years ago. I recently rediscovered the source code and got it running again which was quite a journey sharpoblunto.com/News/2024/02/0…