Research
-
AI and the post-IP internet: what’s worth something when producing another version costs less
Rights are still in force. The question for media companies, creators, and startups is which part of their business depends on controlling a single piece of content, and which part depends on something a new version can’t replace.

-
Ueno Bank: where the crisis stands
A breakdown of what the neobank’s collapse left behind for Latin America’s fintech universe.
-
Vibe-coding and the dream of the agentic CMS
An agent can prepare a page or fix a post. Whether to migrate depends on who reviews those changes, how they reach production, and how much it costs to maintain the site afterward.
-
Is agentic design living up to the expectations?
In a trial with 100 participants, Figma Make sped up interface changes with a defined visual target. Teams already using agents show what it takes to bring those changes to production — and what’s still left to measure.
-
YC Request for Startups Winter 2027: the 11 ideas, opportunities and risks
Y Combinator published 11 requests for its Winter 2027 batch, and only one mentions agents. We analyzed each one with data for and against: from 128-week waits for a transformer to the 2027 Medicaid renewals.
-
Is Wall Street moving to Solana?
JPMorgan, Visa, and Western Union now operate on Solana, but their own money and control still sit on other chains. We break down what’s in production, what’s piloting, and what’s only been announced, and why banks didn’t adopt Solana: they adopted multichain distribution.
-
Between crypto & AI: How can brands survive ecosystem-wide narrative shifts?
From decentralized ownership to machine-generated abundance, the technology industry’s vision of the future has changed dramatically in the past 4 years. How do you build a company that survives the next change of heart?
-
X Marketing Experiment: Does a Premium plan or follower count guarantee reach?
In our Q4 2025 test, a 28-follower account received more views than a 20,000-follower verified account. Follower count, account age and verification did not predict distribution in this sample.






