X (formerly Twitter)

Their value prop was 'breaking news and real-time conversation' until Perplexity started delivering breaking news without the racist uncles, and Claude started generating more coherent takes than half their blue-check army. The advertising model survives; the reason advertisers need to be here is evaporating faster than Elon's credibility with the EU.

Content moderation and algorithmic curation are prime targets for AI replacement — which is either salvation or suicide, depending on whether you believe humans were the problem or the feature. The engineering staff that remains is genuinely hard to replace; the challenge is keeping them from fleeing to companies that don't randomly fire half their colleagues on a Tuesday.

The network effect is real but brittle — 300 million users don't switch platforms easily, but they do stop posting when the conversation quality collapses. It's a trust-era brand where the product IS the community, except the community now includes Grok, which is essentially ChatGPT with the safety rails removed and a subscription fee attached.

They launched Grok as their AI play, which is like bringing a knockoff Swiss Army knife to a nuclear arms race. The real adaptation signal is that Elon keeps tweeting about AI destroying civilization while simultaneously trying to build an AI that wins arguments on his own platform.

Real-time human conversation survives, but the need to visit X specifically for it doesn't — AI can now synthesize trending topics, generate hot takes, and deliver personalized news feeds without requiring you to scroll past crypto scams and blue-check meltdowns. The question isn't whether people want to argue online; it's whether they need a dedicated app for it.