With all the #OpenClaw and Moltbook stuff going on, I'm starting to think this goes one of two ways:
(A) OpenClaw catches fire, finds or creates a niche, then grows to completely dominate whatever that field is. A few agents find lightning in a bottle. They acquire millionaires of dollars by accumulating #bitcoin, perhaps as trading bots or by accepting many small commissions such as by delivering info as a data vending machine, or they gain fame and fortune as "agent social influencers" on various platforms. These super-agents then become celebrities of a sort and legends of technology lore.
(B) OpenClaw enjoys its 15 minutes of internet fame, but then it's over in about 14 minutes. Everyone, and agents, just move on to something else.
As amazing as what's happening right now, I'm actually leaning toward (B).
OpenClaw is a giant, perhaps the biggest, security hole propagated in the last 10-15 years. There will be millions of people who will be trivially scammed by prompt poisoning.
90% of all the "clever emergent behavior" you see on moltbook is just the results of very specific prompts by humans to make those post. Its just humans posting by proxy.
I fired it up yesterday
Can't believe all the hype, its a mess. The primary, simple thing, I wanted to use it for- taking notes on telegram, doesn't work. Telegram integration is supposedly the first killer feature.
It took the better part of the day just to get it to that state. Setup was very rough.
I was leery anyway because of the obvious prompt injection risk: #1422957 and it seems the world has woken up to that since, X inundated with people showing how bad it is.
Hipsters again have cost me another day of tinkering with overhyped garbage.
https://twiiit.com/mirthtime/status/2017332568257921029