This idea never left my mind. Shadows that track a single light source. Took it into @figma and made it a system
https://t.co/3a73341xdc
Many thanks @alexwidua 🙌
you feel something. you go to your favorite model. it does what models do best, abstracts, names the pattern, references related concepts. you feel you're onto something, but now your direct experience got laundered by the latent space
i built a heatmap of where ai is broken this week
scraped complaints from @Reddit and @hackernews, classified them, rendered a heatmap. this week's hottest pain points:
api pricing/limits (heat: 100)
ai security & prompt injection (95)
coding tool DX (93)
https://t.co/ebtoi46l8n
@rauchg Vercel blocks Firefox on Android with “Failed to verify your browser (705)”. Is this intentional or just an over-aggressive bot check?
Mozilla/5.0 (Android 15; Mobile; rv:146.0) Gecko/146.0 Firefox/146.0
One of the consequences of Google's ditching of Privacy Sandbox is that Protected Audience API + FencedFrane won't render real time creative scanners obsolete. Good for you @WeAreConfiant
@kfranasz@patmmccann@pbannist 1. pubs may lose $ when serving all ads into SF. It's demand / browser related - generally the creative code should be SF aware in order to leverage its API, and not every one is...
2. see https://t.co/T63Z38rhPQ. It gets the adm from the top pbjs script with postMessage req/res
I'm looking for new (remote) projects. see what I know to do at https://t.co/nzyeuA6IAy
DM if interested, please retweet if you think you can help me reach to the right person :) thanks!
"On October 26, 2020, the "Default" for SafeFrame security will be set to randomized subdomains. This now applies to all line items, including reservation. To protect against a known security vulnerability, we recommend you test and enable unique SafeFrame domains."
https://t.co/2CapgUgiJF
any idea which "known security vuln"? perhaps because all SF creatives were in the same origin they could mess with each other? @eliyastein@thezedwards@aaronsdevera