Data Rapid fire.
Researchers tested if people could tell the difference between AI and human-generated tweets. Result? They couldn’t. They also couldn’t tell if the tweet was real.
Another one — AI can automate 40% of the average workday, with automation already removing an average of 20% of previously manual tasks. That from a Valor survey, with Salesforce’s saying the same number for automation and adding that More than 50% of sales and service teams don’t know how to get the most value out of generative AI,
The Harris Poll on behalf ofInsight Enterprises finds that within the next three years, most business leaders expect to adopt generative AI to make employees more productive (72%) and enhance customer service (66%).
Around half of Insight’s generative AI report respondents expect to use the technology to assist with research and development (53%) and software development/testing (50%).
Seventy-nine percent of corporate strategists said that technologies such as analytics, artificial intelligence (AI), and automation will be critical to their success over the next two years, according to Gartner, Inc.
Negative backlash — In a Malwarebytes survey, Only 10% surveyed agreed with the following statement: “I trust the information produced by ChatGPT,” while 63% disagreed.
But the one to note — ChatGPT web and mobile traffic fell 9.7% month over month in June. Per SimilarWeb, it’s not just ChatGPT. Character.AI saw 32% less web and mobile traffic in June than the prior month, and Google’s Bard website declined slightly. But also – the top 100 search sites saw a 3.1% dip, and Facebook is down 3.2%.
A critical theory is that students are not using generative AI, so the dip is due to summer – or the broad summer dip in traffic.
I’m a big data nerd, so I noted these data points this week. The story they tell is more important. AI has much potential, and the consensus is that becoming a technology expert will be key. There’s a trust gap, generally found in technologies seemingly appearing out of nowhere, despite their history.
And finally, while one might be tempted to find that usage drop-off worrying, I find it useful. We can link the use to actual user reasons for the product. That’s good news.