“Trained a cost estimator model… emits a probability distribution over multiple buckets… I’m used to training these classifiers by giving it different enums”

The a16z Show · Beyond the God Model | Alex Atallah & Amjad Masad

Play from 40:02
40:02

Audio streams from the publisher’s own feed. Verbatim does not host or copy it.

What was described

Trained an 8B-parameter cost-estimator model that outputs probability buckets (e.g., $5-$10, $10-$20) for Replit prompts

Said nearby on the same subject

“‘…model every 100 days now to a model every 10 days… it was 10 weeks to 10 days actually.’”
AC
Alan Chan
"The Cognitive Revolution" ·
Hard number 100 days → 10 days Share this moment
“The more you post-train models the less diverse their outputs become”
EB
Eric Bigelow
"The Cognitive Revolution" ·
Hard number Share this moment
“Claude wrote me a master prompt ... handed that prompt to Instinct ... Instinct is the middleman between me, Claude and Replit ... I did nothing.”
♫
Said on The Koerner Office - Business Ideas and Deep Dives with Chris Koerner
How they work Share this moment
“The first thing I do is I look at the marketplace… I’ll download probably the top 10 to 20 apps and get really present with them.”
♫
Said on Tony Robbins Podcast Daily
How they work Share this moment

Moments like this, every Monday

The strongest new moments on your topics, in the speakers' own words. Free for three topics.

No card. No password. One-click unsubscribe.

Track a competitor?

Pro watches up to 25 names or phrases and emails you the morning after they are said.

See Pro, $29 a month

Verbatim quotes a short extract and links to the episode; the episode belongs to its publisher. Remove a show.