Watching Fable architect the work, dispatch Opus and Sonnet subagents, and continuously critique and redirect them is the most impressive agentic workflow I’ve used so far.
Honestly, Fable5 felt different.
Most models just take shortcuts to get things done. Fable5 didn’t. It would struggle sometimes, but actually tried to do the work properly.
Sad that we don’t have it anymore. Hope it comes back soon, or later atleast.
Claude and Codex can crush coding tasks, but when it comes to visual comparison—spotting subtle differences, matching layouts, analyzing screenshots—Gemini is in a different league right now. Vision is still an underrated benchmark for AI models.
@DocPriyamMD I was on levofloxacin daily for nearly a year to treat mediastinal TB without any issues. But a year later, just 3 doses of ciprofloxacin for a suspected stomach infection gave me severe brain fog, insomnia, and dizziness—it took almost a 1.5 months to recover.
@Mrsinha The part that really bothers me?
The people suffering most are the poorest Indians breathing this daily.
But instead of solutions we get nationalism + screenshots.
We deserve better conversations than this.
@Mrsinha Comparing with China like it’s a competition is peak whataboutism.
“Someone else is worse today” doesn’t make our air healthier.
Pollution isn’t a cricket match.
@Mrsinha Main causes are well known:
crop burning, insane traffic, construction dust, weak enforcement.
These are OUR problems. Not Western propaganda.