Google’s CEO was seen speaking about a vast array of features that its Gemini AI assistant has been equipped with at this year’s I/O conference. But what he failed to do was address a matter that has been a serious concern for users.
In February of this year, we saw Sundar Pichai pausing the firm’s famous AI chatbot from producing images as many referred to the product as biased. So many users filed complaints against Gemini’s images that featured historical inaccuracies.
It was a serious matter and one that forced Pichai to go public with an apology with promises to fix the issue in a short period. But it’s the middle of May now and there is yet to be any kind of solution in sight.
Moreover, he had reassured users that a couple of weeks was all that was needed to solve the issue, making it appear that it was nothing major. But now that we’re seeing it nowhere in sight, the matter is more than just a small issue.
We saw Google boast so many other features of Gemini that were linked to the yearly I/O developer conference. This entailed customized chatbots to even itinerary planners, not to mention combining with other apps such as Calendar and YouTube Music.
But the matter revolving around producing pictures of people is yet to be turned on throughout its apps online and also across smartphones as mentioned by one of the firm’s representatives.
Now the real question that is on people’s minds is why or what is taking so long? Well, it appears that the issue is more than meets the eye and more complex as well.
The data used for training the model entail more pictures of white individuals than those arising from various other races, cultures, and ethnicities. This limited data has given rise to extreme bias behavior. And it’s high time the company takes the matter seriously and rolls out a better solution so that history isn’t repeated.
Remember, the issue is not a small one and many call it controversial for obvious reasons. Now the real matter of concern is whether or not Google is going to get there and actually amend it or just generate a clumsy quick fix?
Only time can tell whether it reaches that point of interest or not. But whatever the case may be, one thing is for sure. It’s an affair that serves as a serious reminder to the world that fixing AI is never easy, especially when it’s not on its best behavior. And if the root of the issue is biased, don’t underestimate the time required to fix the core issue.
As you can see right now, it’s taking quite some time. And Google knows that. No matter how much it chose to disguise it all by pushing things under the rug, one thing is for sure. It needs attention and the right one to stop criticism from arising again.
Image: DIW / Created Using Bing Image.
Read next: TikTok Creators Unite To Legally Challenge US Government Over New Bill Banning The App
In February of this year, we saw Sundar Pichai pausing the firm’s famous AI chatbot from producing images as many referred to the product as biased. So many users filed complaints against Gemini’s images that featured historical inaccuracies.
It was a serious matter and one that forced Pichai to go public with an apology with promises to fix the issue in a short period. But it’s the middle of May now and there is yet to be any kind of solution in sight.
Moreover, he had reassured users that a couple of weeks was all that was needed to solve the issue, making it appear that it was nothing major. But now that we’re seeing it nowhere in sight, the matter is more than just a small issue.
We saw Google boast so many other features of Gemini that were linked to the yearly I/O developer conference. This entailed customized chatbots to even itinerary planners, not to mention combining with other apps such as Calendar and YouTube Music.
But the matter revolving around producing pictures of people is yet to be turned on throughout its apps online and also across smartphones as mentioned by one of the firm’s representatives.
Now the real question that is on people’s minds is why or what is taking so long? Well, it appears that the issue is more than meets the eye and more complex as well.
The data used for training the model entail more pictures of white individuals than those arising from various other races, cultures, and ethnicities. This limited data has given rise to extreme bias behavior. And it’s high time the company takes the matter seriously and rolls out a better solution so that history isn’t repeated.
Remember, the issue is not a small one and many call it controversial for obvious reasons. Now the real matter of concern is whether or not Google is going to get there and actually amend it or just generate a clumsy quick fix?
Only time can tell whether it reaches that point of interest or not. But whatever the case may be, one thing is for sure. It’s an affair that serves as a serious reminder to the world that fixing AI is never easy, especially when it’s not on its best behavior. And if the root of the issue is biased, don’t underestimate the time required to fix the core issue.
As you can see right now, it’s taking quite some time. And Google knows that. No matter how much it chose to disguise it all by pushing things under the rug, one thing is for sure. It needs attention and the right one to stop criticism from arising again.
Read next: TikTok Creators Unite To Legally Challenge US Government Over New Bill Banning The App
