DeepMind Leading to a Deep Failure or Deep State of AI
Google's failure in AI is the end of an era for Silicon Valley
Artificial Intelligence has become a weapon of profit in the era of surveillance capitalism and the corporate will for supremacy. So the best AI should theoretically win most of the time right?
While DeepMind extends its hunt for AI talent to Toronto, its bid for autonomy from the mothership is not going so well. DeepMind’s profitability is also not going so well.
DeepMind, the London-based artificial intelligence lab acquired by Google in 2014, is even making outrageous headline claims about the future of AGI. Reinforcement learning enough for artificial general intelligence? In its quest for artificial general intelligence, which is sometimes called human-level AI, DeepMind is focusing a large chunk of its efforts on an approach called “reinforcement learning.”
Sounds about as realistic as how Waymo made promises it didn’t keep and nearly the entire C-level suite was let go.
A Deep Failure at Google
The lack of AI regulation and ethics and the pressure to monetize faster than the other guy is creating a dangerous wild-wild west scenario of AI firms and simply companies with deeper pockets using AI for their R&D benefit.
Curiously the AI leaders of the world are also some of the least ethical like Google, ByteDance, Amazon and so forth. They will do anything to win, just for the sake of profit and global control over their ecosystem and industry.
Computer scientists are questioning whether DeepMind, the Alphabet-owned U.K. firm that’s widely regarded as one of the world’s premier AI labs, will ever be able to make machines with the kind of “general” intelligence seen in humans and animals. Alphabet’s hyper expensive AI lab is mostly only good at creating AI hype and headlines for the mothership.
The Least Profitable AI Lab in the World
Just how unprofitable is DeepMind? DeepMind A.I. unit lost $649 million in 2019 and had a $1.5 billion debt waived by Alphabet. It likely lost a lot more money in 2020. Even as Google keeps messing up AI ethics and inclusion in its AI practices.
DeepMind, which was acquired by Google in 2014 for around $600 million, believes that AI systems underpinned by reinforcement learning could theoretically grow and learn so much that they break the theoretical barrier to AGI without any new technological developments.
Alphabet has a habit of funding “other bets” that for the most part make no sense and even AI academics disagree with what DeepMind is saying. This is what happens when a firm is near failure, the false promises begin to sound even a bit absurd. While Alphabet can continue to fund AI especially now that it has Google Cloud, advertising and cloud are the two best cash cows in technology, Google’s ethical direction is highly questionable.
Google could lead in chip design perhaps to save DeepMind. Google’s new AI can draw up a chip’s “floorplan.” This essentially involves plotting where components like CPUs, GPUs, and memory are placed on the silicon die in relation to one another — their positioning on these miniscule boards is important as it affects the chip’s power consumption and processing speed.
Maybe Bloated AI Labs Are Not the Answer to AGI
Researchers at the company, which has grown to around 1,000 people under Alphabet’s ownership have battled for their quasi-independence and are now part of Alphabet’s sketchy monetization plans for artificial intelligence. Like scientists at the mercy of the big bosses in a cartoon, we might be witnessing the gradual decline of DeepMind.
While OpenAI forges partnership with Microsoft’s friends and minions, China’s Wu Dao is showing a sense of scale and progress. DeepMind is a great AI research firm, and decent at great at creating headline buzz, but hasn’t made a substantial breakthrough in years either in AI or for Google.
DeepMind is more than double the size of Facebook AI Research (FAIR) with well over 1,000 employees. The London-headquartered AI lab had a loss of £477 million ($649 million) in 2019, up from £470 million in 2018, according to documents filed with the U.K.’s Companies House registry. The vast majority of DeepMind’s spending in 2019 went on “staff and other related costs.”
In Google’s bid to own the internet with its search and advertising monopoly, its AI is not doing so great. AlphaGo might go down as DeepMind’s most significant achievement. The reality is today DeepMind is in crisis both internally, in relation to Google and in its inability to monetize and make practical its expensive R&D and AI researchers making huge salaries.
What’s worse is if DeepMind were to succeed in its goals, it would make Alphabet in the deep state mother-AI of surveillance capitalism.
You can read more articles like this at the Last Futurist.