[ad_1]
“We are teaching machines how to think.” That’s a quote from the recent video posted by Joe Toscano, a former Google consultant. In a recent video posted on Fox Business, Toscano talked about how Google went “too big too soon” with AI, and this is a sentiment that we all share.
Now, Google is dealing with a bit of a mess with its AI products, and it’s been that way ever since the launch of Bard. While Gemini and other Google AI products are useful and functional, there is a deeper issue.
A former Google consultant says that the company went “too big too soon” with AI
In the two-and-a-half-minute-long video, Toscano stated that a lot of historical inaccuracies produced by chatbots are a product of Google and other companies just feeding the models information without really explaining what’s correct and what’s not correct. So, this is one reason why they are liable to give you inaccurate details. It’s “likely that [Google] fed it a bunch of data and, in the modern machine learning environment we’re in, they expect that the machine will learn on their own and that the outcome will be, more or less, bulletproof.”
Toscana also explained why Google is struggling with Gemini. He explained the major issue of the company playing corporate catch-up. Google Bard was not launched as a product, it was launched as a response. When ChatGPT launched, it threatened Google immensely. So, the company had to rush to get something to the market in order to keep the investors happy.
Google, like other companies, is trying to build a be-all-end-all artificially intelligent system that is as smart as a human being. However, the circumstances just aren’t conducive for that in Google’s case. Google is a major profit-driven corporate company that needs to justify every cent it’s spending on AI development. So, aside from developing AI, the company also has to worry about profits and investors. So, it has to keep up with the times and keep launching new AI products to compete with companies such as OpenAI, xAI, Anthropic, Etc.
Corporate catch-up
Because of this, Google has had to rush products onto the market without proper testing. The most recent example is the Gemini image generating people with historically and accurate skin tones, ethnicities, and genders. Another example is how half-assed Gemini is as an assistant.
It seems that all of Google’s AI products are just being thrown onto the market, and this creates a bad track record. In the video, Toscano states that this is what happens when companies play corporate catch-up. Companies don’t properly test their products because their profits and stock prices are at stake. They put it on the market and hope that it sticks.
If ChatGPT did not launch when it did, and Google had the time to properly flesh out its AI products, then things would be different. However, it has to compete with ever-evolving competition. For example, OpenAI is testing Sora. Well, we can bet that Google is also testing its own video creator. However, since it’s already behind the times, we can bet that it won’t be as polished.
This is why Apple’s AI might be better
News and rumors are already ramping up about Apple’s upcoming AI product. While Apple is new to the game in terms of AI, we can comfortably assume that, when Apple does release it, it’ll be much more refined and better integrated than Google’s ecosystem of AI products.
The thing about Apple is that the company doesn’t typically launch products until they pretty much work perfectly. Apple may be able to integrate its AI flawlessly with iOS, iPadOS, and macOS. We can trust that it’s going to take the proper time to refine its experience because it’s not at risk of losing ad revenue like Google is, and the vast majority of its income comes from hardware. So, it’s not much reason to rush this product to the market.
[ad_2]
Source link