New research from the University of Irvine is shedding light on some very interesting findings regarding bots and their intelligence.
You may have come across CAPTCHAs plenty of times when you’re online. They’re usually thought of as the perfect test that’s taken when you plug in your password across a particular site. Moreover, the test always mentions that the site may wish to confirm that you’re a human and not a robot so that’s the justification provided for embarking on the ordeal.
But you’ll be amazed at how the latest research on this front is showcasing how these kinds of tests aren’t really too effective. In fact, they not only take up a user’s time but bots end up solving them at a faster pace than any other human out there. Did we mention how they’re doing it with greater accuracy than most of us combined?
The automated tests dubbed CAPTCHAs were in the spotlight in this new study. And that’s when the authors were busy comparing close to 1400 members of the public and a similar figure for bots across similar tests too. And you’ll be shocked to learn how bots continued to overcome the capabilities and talents of the human race by miles.
For example, humans were requested to solve some very disoriented CAPTCHAs. And the humans ended up doing it in less than 15 seconds which really does sound too good to be true. But wait, the reality is far from that. And that’s because most of the answers that were generated were incorrect. And we’re talking close to 50 to 84%.
But when the bots were told to do the same, they did it in time spans that lasted for less than one second. And shockingly, the accuracy was close to 99.8%. Moreover, a recent study by The Sunday Times shed light on how CAPTCHAs were already being solved so well by bots but it lacked a decent comparison with humans, which this new research managed to do.
This research’s measurements went on to add in a confident manner how so many humans were much slower in pace than bots and it’s proven. It’s a result that was never proven in the past and we’re still waiting for the results of the research to be peer-reviewed, things are looking great so far.
So this study is also hinting at how better strategies must be explored because if this CAPTCHA technology can be solved by bots faster and better then why are we all still using it? Why aren’t more efforts being done to establish better means for verification? Above all, this technology is disliked all over the globe and it certainly does not come cheap either. Did we mention how much time it wastes us, humans, too?
Being ineffective against bots is never going to be appreciated. And that’s the biggest reason to call for change, we feel. But despite knowing all of this, so many huge firms are reluctant in terms of admitting that such tests are poor in design. After all, they’re making some handsome funds through sales of them, despite knowing their major lackings.
So many firms aren’t willing to accept that these tests are not worth it anymore. They’re continuously used for webpages and so many of them too. Despite being majorly ineffective, the rise of these kinds of studies is clear proof that we need to find options that go beyond counting the number of stairs or searching for traffic lights in the most unclear images. And that is being done to prove humans aren’t bots, despite them beating us a long time back.
Read next: AI Will Disrupt Most Industries In The Near Future, Databricks CEO Boldly Confirms
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment