HN user

tabletcorry

1,600 karma

Some random security engineer, more at <sharpletters.net>

Posts22
Comments12
View on HN
alexgaynor.net 7d ago

You can't bug fix your way out of the vulnpocalypse

tabletcorry
5pts0
www.theverge.com 1mo ago

The Steam Machine is the most ambitious game console I've ever played

tabletcorry
4pts0
openai.com 1mo ago

Daybreak: Tools for securing every organization in the world

tabletcorry
19pts1
www.theverge.com 8mo ago

Signal's secure message backups arrive on iOS

tabletcorry
3pts0
techcrunch.com 8mo ago

ChatGPT launches group chats globally

tabletcorry
2pts1
opensource.microsoft.com 8mo ago

Microsoft makes Zork open-source

tabletcorry
640pts242
openai.com 8mo ago

Early experiments in accelerating science with GPT-5

tabletcorry
18pts0
fly.io 8mo ago

You should write an agent

tabletcorry
1070pts395
www.theverge.com 8mo ago

Epic and Google agree to settle their lawsuit and change Android's fate globally

tabletcorry
10pts2
tailscale.com 8mo ago

Better authentication with workload identity federation

tabletcorry
4pts0
tailscale.com 8mo ago

App capabilities, now for all your apps

tabletcorry
2pts0
fly.io 9mo ago

Kurt Got Got

tabletcorry
353pts241
magoo.medium.com 1y ago

The "Working from China" Problem

tabletcorry
3pts0
www.anthropic.com 1y ago

Claude Gov Models for U.S. National Security Customers

tabletcorry
43pts4
openai.com 1y ago

Disrupting malicious uses of AI

tabletcorry
1pts0
mjg59.dreamwidth.org 1y ago

Twitter's new encrypted DMs aren't better than the old ones

tabletcorry
228pts205
www.youtube.com 1y ago

Making the most powerful Red Bull

tabletcorry
2pts0
fly.io 1y ago

My AI skeptic friends are all nuts

tabletcorry
2356pts2826
sharpletters.net 1y ago

Pick Your (User Agent) Battles

tabletcorry
2pts0
help.openai.com 1y ago

Complimentary tokens on prompts and completions shared with OpenAI

tabletcorry
1pts0
sharpletters.net 1y ago

HDR‑Infused Emoji

tabletcorry
274pts146
fly.io 3y ago

API Tokens: A Tedious Survey

tabletcorry
3pts0
GPT-5.2 7 months ago

The matrix required for a fair comparison is getting too complicated, since you have to compare chat/thinking/pro against an array of Anthropic and Google models.

But they publish all the same numbers, so you can make the full comparison yourself, if you want to.

GPT-5.2 7 months ago

Slight increase in model cost, but looks like benefits across the board to match.

  gpt-5.2 $1.75 $0.175 $14.00
  gpt-5.1 $1.25 $0.125 $10.00

This level of knowledge is nearly impossible to maintain as the codebase grows though, beyond one or two people at a typical company. And tools need to exist for the new hire as well as the long-standing employee.