355
Audio & Video Production344
Automation & Workflow224
Software Development250
Marketing & Growth192
AI Infrastructure & MLOps174
Writing & Content Creation203
Data & Analytics140
Design & Creative169
Customer Support131
Photography & Imaging156
Sales & Outreach125
Voice & Speech135
Education & Learning131
Operations & Admin87
A new discussion highlights why AI can save time on tasks but still fail to improve company results, due to extra checking, switching tools, and unclear data.
In short: Many workers say AI saves them time, but much of that time is spent managing AI, so companies may not see better overall results.
Workplace leaders often talk about AI boosting productivity, meaning people get more done in less time. A Financial Times newsletter highlights a problem: saving minutes on small tasks does not always make an organisation run better.
One issue is “botsitting”, which means people have to babysit AI outputs (like proofreading a rough draft and fixing errors). Rebecca Hinds at Glean’s Work AI Institute said workers report spending 6.4 hours a week on this hidden work, such as giving the AI more background, checking answers, and rerunning requests.
A separate issue is the “toggle tax”. This is time lost switching between different AI tools, like having to visit several shops to buy one item. The survey of 6,000 digital workers found nearly 8 in 10 people juggle multiple AI tools each week, and 60% run the same question in different tools to get a better result.
The survey also found a gap between time saved and business impact. It reported that AI saves workers 11 hours a week, but only 13% of workers see any improvement in company performance.
Several leaders quoted in the piece argue that companies may need to fix basic issues first, like having organised data and clear processes, before AI can help at a company level. It may also mean redesigning teams and workflows so AI fits naturally, instead of being placed on top of old ways of working.
Source: Financial Times