When AI began to demonstrate that it could be competent at writing code, an entire industry shook. Software engineers seemed like one of those jobs that was going to be quickly eclipsed by the arrival of AI, but It turns out that in the end it was the other way around.
The employees are still there, the CEOs of companies that are very optimistic about the future of AI say that, precisely, more engineers will be needed in software that never and AI is compressing work that takes hours to just seconds. But there is a problem: this is not translating into shorter days, quite the opposite.
And all of this is due to a management problem that forces workers to jump between projects haphazardly.
A management problem, not a technology problem
In a recent articleBusiness Insider exposes how AI has transformed the routine of six workers at technology companies such as Amazon. In the interview, They detailed that using AI is saving them a lot of time And the most interesting thing: they quantify that work saved.
According to their statements, the use of AI to summarize meetings, review code, automate reports that they have to do frequently and write documents has allowed them to save a day of work each week, which is said to be early. Does that mean they have an extra day of rest a week? Obviously… no. In fact, some work more hours than before.
In these statements, one of the Amazon employees stated that this time saved is not to have a couple of coffees watching videos to clear the mind, but rather automatically redirects to other projects. Another of the engineers commented that building these automation systems is adding more work hours to his week, since it is also You have to review those processes constantly.
The BI article is very limited because the sources are scarce, but a Boston Consulting Group report called Global AI in which 12,000 employees of leading technology companies were interviewed states that 42% saved the equivalent of one day of work per week, but 66% said they have no idea what to do with that “recovered” time.
Not rest, of course, but go to other projects for which they do not have any type of management. It’s as if I had to do three articles a day and then, since I have time to do them with the AI, I start editing a video to help the video department because… well, because I have time and I have to do something.
Global AI is not the only company with a larger sample. Another study in which interviewed survey of 3,200 business leaders found that 85% of employees They save between one and seven hours of work a week thanks to new tools, but almost 40% of that recovered time is immediately lost in reviewing, correcting and partially or totally redoing those results generated by AI. It doesn’t make any kind of sense.
“Please don’t use AI just for the sake of it” – Dave Treadwell, Senior Vice President at Amazon, to his team
Technology companies are already pointing out that there is a huge management problem. Faced with the unbridled optimism of some and the “we have to use it just because, because now we are an AI company” (Goal, for example, creating competitions to see who uses it the most), there is the other side of the coin.
There are already bosses who are pointing that managers continue to be obsessed with the workforce instead of rethinking workflows, as well as voices that suggest that AI not be used just for the sake of using it.
As you can see, there are many studies that point out that, indeed, AI is saving time in certain jobs, but all that time is wasted because no one has instructions on what to do with it.
In The Next Web They did some research on this and the conclusion was the same: a tool that is capable of saving an employee an hour is only as useful as the company’s ability to do something with that hour. And, according to the article, “the evidence so far indicates that most are not doing so.”
In fact, this same week, another 404 report in which they detailed how Google’s own engineers who write the AI code they laugh at Google’s AI He stressed that there is a disconnection between work policies and the need to push the development of AI.
“We are finding that AI has relieved the pressure and bottleneck in code generation,” said one employee, “but everything else has become a bottleneck: build times, testing, the delays in human review, the comparatively slow infrastructure, and the version comparison system.”
GIPHY App Key not set. Please check settings