dmesg --follow

[ 66151680.000 ] posts.x: Full talk on training a search sub-agent with RL, and the speed and cost gains that come with it: | [ 66153540.000 ] posts.x: @hypeapps This kind of time tracking is quite important. Until you’ve logged how long projects (and, if you’re as easily distracted as I am… | [ 66153720.000 ] posts.x: @Mappletons I’ve been redesigning my personal site this past week and I’ve studied your site extensively as a great example. I really appreciate your… | [ 66153780.000 ] posts.x: @Mappletons The redesign isn’t live yet, BTW … | [ 66159420.000 ] posts.x: Full talk on why there's no single right chunk size and how multiscale indexing with RRF closes the gap: | [ 66159420.000 ] posts.x: A fixed chunk size is a bet on queries you haven't seen yet. @yuvalinthedeep, Sr. Developer Advocate at AI21, tests that bet in "Stop Chunking Like… | [ 66166140.000 ] posts.x: @hypeapps @toggl Is this because you’re straddling multiple agent sessions for multiple projects simultaneously? | [ 66166560.000 ] posts.x: @hypeapps @toggl You could still get pretty close. Agent sessions are stored on disk. Subtract the long gaps between your inputs to the session… | [ 66168660.000 ] posts.x: Full talk on why llms.txt isn't enough and what actually makes a website agent-ready: | [ 66168660.000 ] posts.x: Almost half of the websites in one study already publish an llms.txt file for agents to read, but almost none of the agents actually use it… | [ 66169620.000 ] posts.x: Full talk on why AI cluster networks need a receiver-driven, message-based protocol instead of TCP: | [ 66169620.000 ] posts.x: Most AI clusters still tune their networks for giant weight transfers, but the workloads pushing performance limits now are tiny messages: a KV cache… | [ 66179400.000 ] posts.x: Full talk on the harness layers, the files-vs-databases tradeoff, and context rot: | [ 66179400.000 ] posts.x: Most of what makes an AI agent reliable has nothing to do with the model itself. In "Total Recall: Agent Memory and Harness Engineering,"… |

[ 66151680.000 ] posts.x: Full talk on training a search sub-agent with RL, and the speed and cost gains that come with it: | [ 66153540.000 ] posts.x: @hypeapps This kind of time tracking is quite important. Until you’ve logged how long projects (and, if you’re as easily distracted as I am… | [ 66153720.000 ] posts.x: @Mappletons I’ve been redesigning my personal site this past week and I’ve studied your site extensively as a great example. I really appreciate your… | [ 66153780.000 ] posts.x: @Mappletons The redesign isn’t live yet, BTW … | [ 66159420.000 ] posts.x: Full talk on why there's no single right chunk size and how multiscale indexing with RRF closes the gap: | [ 66159420.000 ] posts.x: A fixed chunk size is a bet on queries you haven't seen yet. @yuvalinthedeep, Sr. Developer Advocate at AI21, tests that bet in "Stop Chunking Like… | [ 66166140.000 ] posts.x: @hypeapps @toggl Is this because you’re straddling multiple agent sessions for multiple projects simultaneously? | [ 66166560.000 ] posts.x: @hypeapps @toggl You could still get pretty close. Agent sessions are stored on disk. Subtract the long gaps between your inputs to the session… | [ 66168660.000 ] posts.x: Full talk on why llms.txt isn't enough and what actually makes a website agent-ready: | [ 66168660.000 ] posts.x: Almost half of the websites in one study already publish an llms.txt file for agents to read, but almost none of the agents actually use it… | [ 66169620.000 ] posts.x: Full talk on why AI cluster networks need a receiver-driven, message-based protocol instead of TCP: | [ 66169620.000 ] posts.x: Most AI clusters still tune their networks for giant weight transfers, but the workloads pushing performance limits now are tiny messages: a KV cache… | [ 66179400.000 ] posts.x: Full talk on the harness layers, the files-vs-databases tradeoff, and context rot: | [ 66179400.000 ] posts.x: Most of what makes an AI agent reliable has nothing to do with the model itself. In "Total Recall: Agent Memory and Harness Engineering,"… |

~/pages/about.md -rw-r--r-- 114 words 2026-09-29

Heya! I'm Corey.

I build AI-native businesses at Rexmore. I'm also an artisan of specialty coffee and a pickleball fanatic.

I serve on the boards of companies and non-profit organizations. I'm a frequent speaker and writer on topics in artificial intelligence, technology and business.

I did other stuffs before this:

Founding faculty and Adjunct Professor of Graduate Machine Learning at a university in Chicago
Chief Innovation Officer of the commercial technology business at a large consulting firm
Partner and Head of Finance at a Wall Street FinTech company

I share insights from the frontiers of AI engineering. You should subscribe to my newsletter. It's an email sent when I have something worthwhile sending. Replies are encouraged!

corey@gallon.me:~$ tail -f /writing Attach to the stream. An email when I have something worth sending. Replies encouraged!