- Apr 4, 2009
- 13,911
- 71,109
Source: Swedish Television interview [Video]
Google is used to being unchallenged at the top of the search results. But then the AI service ChatGPT came along last fall and radically changed the playing field. In an interview with SVT, Google's CEO Sundar Pichai says that competitors' entry into the market has caused them to launch services that they might have preferred to wait with:
- We would probably have given it another year before we launched.
Over 90 percent of all searches in the world usually go through Google, which has brought in hundreds of billions of dollars every year. Then came the newcomer ChatGPT and gained 100 million users in two months, who could now search for ready-made answers instead of "Googling" for links. How does it feel to be run over like this?
- It's like when Alexa and Siri came (Amazon's and Apple's voice-based AI assistants), then we got the same questions. The important thing for us is not to be first, but to get it right, Google and parent company Alphabet CEO Sundar Pichai tells SVT in an exclusive interview when he is on a temporary visit to Stockholm.
Pressured to launch
There was no need to be second on the ball. The technology behind these language models that ChatGPT and Bard use was partially developed by Google in 2017. Why didn't they themselves launch an AI assistant like this earlier?
- I have been focused on developing the systems in a safe and responsible way. They have downsides, they are not always reliable. As we are focused on delivering credible information, we were more cautious in using these models.
But when they were overtaken by ChatGPT in November, the full focus was on boosting their own services with as much AI as possible. The own chatbot Bard was recently launched. But it is clear that they would have preferred to let it take a little more time.
- We would probably have given it another year before we launched, admits Pichai.
What risks does Google see in using chatbots today?
- If you come to us to seek information about the correct dosage of a medicine, then you want it to be correct. It is an example of a question that these models are not yet ready to answer correctly.
Sundar Pichai started the day by meeting Prime Minister Ulf Kristersson, among other things, to talk about AI. Sweden holds the presidency of the EU, which is developing regulation of artificial intelligence.
- AI is too important not to be regulated, it is also too important not to be regulated well. We want to avoid considering all applications of AI as high risk. This is a technology that touches every industry and an opportunity for Europe to embrace innovation that can strengthen competitiveness.
Google is used to being unchallenged at the top of the search results. But then the AI service ChatGPT came along last fall and radically changed the playing field. In an interview with SVT, Google's CEO Sundar Pichai says that competitors' entry into the market has caused them to launch services that they might have preferred to wait with:
- We would probably have given it another year before we launched.
Over 90 percent of all searches in the world usually go through Google, which has brought in hundreds of billions of dollars every year. Then came the newcomer ChatGPT and gained 100 million users in two months, who could now search for ready-made answers instead of "Googling" for links. How does it feel to be run over like this?
- It's like when Alexa and Siri came (Amazon's and Apple's voice-based AI assistants), then we got the same questions. The important thing for us is not to be first, but to get it right, Google and parent company Alphabet CEO Sundar Pichai tells SVT in an exclusive interview when he is on a temporary visit to Stockholm.
Pressured to launch
There was no need to be second on the ball. The technology behind these language models that ChatGPT and Bard use was partially developed by Google in 2017. Why didn't they themselves launch an AI assistant like this earlier?
- I have been focused on developing the systems in a safe and responsible way. They have downsides, they are not always reliable. As we are focused on delivering credible information, we were more cautious in using these models.
But when they were overtaken by ChatGPT in November, the full focus was on boosting their own services with as much AI as possible. The own chatbot Bard was recently launched. But it is clear that they would have preferred to let it take a little more time.
- We would probably have given it another year before we launched, admits Pichai.
What risks does Google see in using chatbots today?
- If you come to us to seek information about the correct dosage of a medicine, then you want it to be correct. It is an example of a question that these models are not yet ready to answer correctly.
Sundar Pichai started the day by meeting Prime Minister Ulf Kristersson, among other things, to talk about AI. Sweden holds the presidency of the EU, which is developing regulation of artificial intelligence.
- AI is too important not to be regulated, it is also too important not to be regulated well. We want to avoid considering all applications of AI as high risk. This is a technology that touches every industry and an opportunity for Europe to embrace innovation that can strengthen competitiveness.
