Kleine Unternehmen sind noch nicht ganz von der künstlichen Intelligenz überzeugt – Berichten zufolge nimmt die KI-Einführung bei kleinen Unternehmen stark zu. Aber diese wahrscheinlich überzogenen Behauptungen tragen dazu bei, ihre Marken zu stärken
https://www.theguardian.com/business/2025/jan/12/small-businesses-ai
3 Comments
From the article
>[Artificial intelligence](https://www.theguardian.com/technology/artificialintelligenceai) (AI) adoption is “[surging](https://www.verizon.com/about/news/verizon-business-state-of-small-business-survey-ai)” at small businesses, according to Verizon. Salesforce [says](https://www.salesforce.com/news/stories/smbs-ai-trends-2025/) AI is driving “stronger revenue growth” at small businesses. The UD Chamber of Commerce [reports](https://www.uschamber.com/technology/small-businesses-are-leveraging-ai-to-compete-and-thrive) that businesses are “leveraging AI to compete and thrive” and “almost all small businesses are using a software tool that is enabled by AI”, the Associated Press [claims](https://apnews.com/article/small-business-artificial-intelligence-productivity-f6fa7b2a1ce0a9f2e5b8b48670b3098a).
I truly don’t understand why posteruing around AI is creating interest or confidence for investors. I understand that at the core level, using AI should indicate a certain level of efficiency, but short term efficiency gains shouldn’t be more attractive than it is unattractive to create a reliance on unproven and unoptimized technology. For example, the street seems to praise Meta for announcing that AI would replace Engineers, but at this current point in AI that would be like getting on a plane and hearing that the pilots are actually high school interns who have only ever practiced on Microsoft Flight Sim. I’m getting off the plane.
I think the real big shift will be when something like replit has its ‘ChatGPT’ moment when all of a sudden the things AI couldn’t do in regard to coding a whole app or site comes together and it can. I’m a real believer that this is not only inevitable but probably only a couple years away. There’s only so many ways to code the majority of applications and plenty of resources, source code, and repository practices to review to figure it out after throwing enough money at it. I think you’ll still need plenty of technical people at a certain level, but an MVP could be done with minimal staffing and just ‘Development as a Service’ and a few other ‘as a Services’.
That will take potential small business people from needing to consider giving substantial equity out or borrowing millions of dollars for developers to paying a few thousand dollars or maybe tens of thousands to get an MVP made and then they can just scale up as the product gains traction.