embhow

@scribbledawn.bsky.social

100 Followers 209 Following

impact interactive wordjam heartune 🇨🇦->🏴󠁧󠁢󠁳󠁣󠁴󠁿->?? Prism: https://eliot-mb-howard.itch.io/prism-a-tale-of-heterogenous-futures Game Director Story (demo): https://store.steampowered.com/app/1305180/Game_Director_Story/ eliotmbhoward.com

street art outside a café in which an adorable cartoon corgi pointing to a line-goes-up graph???

A chart on the goals of AI Alignment, from "Instructions" ("the agent does what I instruct it to do") to "Values" ("the agent does what it morally ought to do, as defined by the individual or society"). Includes limitations/risks for each entry. Here's the risks for "values": "current the best possibility, but it still encounters two difficulties of 1) specifying what values or principles, and 2) concerning the body of people who select the principles with which AI aligns.

A screenshot of Loris.ai's own website, promoting the subheadings: "ingestion of omnichannel conversations", "AI models analyze conversations", "Generate conversation intelligence", "Surface findings to your team"

Featured on the left is a "Sentiment Graph" featuring Agent Insights: "Soft skills up 93%! Detection up 89%, Solution up 83%, Closure down 48%."

"...If our group can’t solve this societal problem on their own, can they at least practice advocacy for the individuals who fall under their purview?

And that. There is the closest thing to an ideal I have. It’s twisty, squirrely, I have to work to maintain it always, but it comes down to this: kindness, defiant of context. We can’t ignore the fact that the weight of history compresses all our available choices; all we can do is practice moments which demand the best from us, so we might show up prepared. "