Google Will Invest Over US$ 100 Billion in Artificial Intelligence, Says CEO of DeepMind, Highlighting the Company’s Computational Advantage and the Technology’s Potential for a New Scientific Renaissance
Google is preparing one of the largest investments in its history, allocating an impressive US$ 100 billion for the development and advancement of artificial intelligence.
Demis Hassabis, CEO of Google DeepMind, revealed at a TED conference in Vancouver that Google plans to invest over US$ 100 billion over time in the development of artificial intelligence (AI).
During the presentation, he discussed the company’s ambitious plans regarding AI technology, highlighting that the investment would be even greater than that projected by OpenAI and Microsoft for building a supercomputer in a data center called “Stargate,” at up to US$ 100 billion.
-
NASA will launch a new space telescope in September that will do in just one year what the Hubble would take 2,000 years to complete, and its field of view is 100 times larger than the James Webb’s, creating the largest atlas of the universe ever made.
-
São Paulo will install giant screens up to 1,000 square meters at the city’s most famous intersection, inspired by Times Square, and the largest of them will cover the entire facade of a building.
-
German scientists create colorful solar panels with record efficiency inspired by butterfly wings
-
Is this the end of the 5G network? Chinese take the lead in technology and are already starting tests of the first “Pre-6G” network this week – the technology is in the testing phase and promises to be a revolution in the sector.
In response to a question about the “Stargate” project, Hassabis stated that Google is investing more than previously disclosed estimates, without providing details on the exact figures.
Google Does Not Want to Fall Behind in Artificial Intelligence
The statement came after Bloomberg released information regarding the plans of OpenAI and Microsoft, which involve a supercomputer with millions of AI chips to enhance large-scale computing.
Hassabis, who founded DeepMind in 2010 before its acquisition by Google in 2014, also spoke about Alphabet’s competitive advantage, Google’s parent company.
He emphasized that the company has superior computing power compared to its rivals, such as Microsoft, which he considers essential to achieve Artificial General Intelligence (AGI).
AGI is the stage at which AI achieves human-like intelligence across various tasks, something Google, according to him, is closer to achieving due to its vast computational power.
In previous statements, Hassabis compared the growing hype around AI to the hype that surrounded cryptocurrencies.
In March, he stated that investment in AI “comes with a lot of hype and perhaps some pitfalls,” and that it distracts from the true science and research behind the technology.
Despite this, he expressed optimism about the future, saying that the industry is “just scratching the surface” of what is possible, and envisioned the beginning of a new era of scientific discoveries, comparing it to a “new Renaissance.”
New Google Chip
Google has taken a significant step in quantum technology with the launch of Willow, its latest quantum computing chip. Designed to tackle the most complex challenges in the field, Willow promises to redefine the performance and efficiency of quantum computing.
Key Features of Willow
- Expanded Qubit Count:
Willow incorporates 105 qubits, representing a significant advancement compared to its predecessor, Sycamore, which had 72 qubits. - Error Correction Innovation:
One of Willow’s key features is its ability to reduce errors exponentially as the number of qubits increases. Due to the sensitivity of qubits to external factors, error correction is a significant challenge in quantum computing. - Superior Performance:
During benchmark tests, Willow performed complex calculations in less than five minutes, a task that would take about 10 septillion years to complete by today’s most powerful supercomputers. - Advanced Logical Qubits:
Willow adopts an innovative approach by grouping multiple physical qubits to form logical qubits, resulting in significantly lower error rates. While individual qubits have high chances of error, Willow’s logical qubits demonstrate an error rate of only 1 in 1,000 per computation cycle.

-
1 person reacted to this.