Nitter's Resilience vs. IBM's Bob: Platform Control Gets Real
Nitter, the Twitter/X frontend that was largely killed off by API restrictions and targeted takedowns, now has more working instances than before those takedowns. The Streisand effect is working. HN commenters are split between celebrating the resilience of open alternatives and noting that links to Nitter instances tend to rot, making them a liability for long-term references.
In the same news cycle, IBM launched 'Bob,' an AI assistant that immediately drew comparisons to Microsoft Bob and was described by people who used it as absolutely dreadful. The product is named after a legendary interface failure, which is either brave or oblivious.
The through-line: platforms that try to kill off access for third parties (X blocking Nitter, Anthropic blocking OpenClaw) tend to generate resilient workarounds, while companies building AI products with no clear user value create noise that drowns out signal.
So what?
Founders building on top of platforms with restrictive API policies should treat access as a product risk, not a stable input. The Nitter story shows that communities will route around restrictions, but it also shows that those workarounds are brittle and cannot be built into a product roadmap responsibly.