has anyone noticed a sharp decline in incremental reach, last week with the trend continuing this week? nothing noteworthy changed, this account has sunk from 35-40% to <10%
@PhilKiel know you keep a close eye on this stuff, everything stable your end?
can confirm rumours, fable is @AnthropicAI 's most capable model yet. solved a 6-month bug in 1.5hrs this morning. can also confirm it used a record amount of tokens for one session. pick your model wisely.
@TaylorHoliday@BryanBumgardner@northbeam agree this video isnt necessarily demonstrating poor performance. very plausible the delivery system is better balancing exploration with exploitation. would be keen to see 1st-mer as the x-axis.
contrary to what tempo suggests, i'm looking to hire a graphic designer. ideal candidate is obsessed with static ads, ai-native (tooling) and performance driven.
workday: mornings building your swipe file; afternoons producing. every day ends with new statics ready to test.
fully remote. contractors, part-time and full-time considered.
industry: adventure travel
reshare or tag someone built for this.
dms open.
holy moly this. have over 300 books with highlights on the backlog. i created an app which took screenshots of pages and used ocr to strip highlighted passages and dump them to storage for enrichment/obsidian but its so cumbersome. about to happily sunset if this thing actually works. thanks for the share.
im transparently naive to the bidding/placement model. i listen to a podcast and the host has obviously managed the injection of advertising, is that still true? if so, surely a better model is to dynamically insert ads on a user level and split the income? it would allow for much more personalization and dynamism, which in turn would simultaneously improve the user experience and performance
@jacob_posel i know you know, so lets get straight to it. what's the advantage of hq vs a company that has been educated to push prs and the rest is thoughtfully managed?