Google DeepMind Workers Urge Company to Drop Military Contracts - 1

Photo by Adarsh Chauhan on Unsplash

Google DeepMind Workers Urge Company to Drop Military Contracts

  • Written by Andrea Miliani Former Tech News Expert
  • Fact-Checked by

Around 200 Google DeepMind workers signed a letter urging the tech giant to drop military contracts with multiple organizations earlier this year. The information has been revealed by TIME magazine in an exclusive with more details of the dispute.

TIME had access to the internal document, dated May 16th, and confirmed that multiple talents in the AI labs are “concerned by recent reports of Google’s contracts with military organizations.” The document also clarifies that it is not addressed to any particular conflict or current geopolitical situation.

In April , TIME revealed that Google has a contract with the Israeli Ministry of Defense for a program called Project Nimbus to supply AI and cloud computing services.

“Any involvement with military and weapon manufacturing impacts our position as leaders in ethical and responsible AI, and goes against our mission statement and stated AI Principles,” wrote the workers in the letter.

The machine learning talents that signed the petition represent only 5% of Google’s DeepMind workforce but reveal important internal differences and approaches towards how Google’s AI technology has been evolving.

According to the magazine, Google acquired DeepMind in 2014 and it had kept its autonomy and original principles until recent years. In 2021 the lab leaders tried to regain autonomy to get a more independent legal structure but the request was denied , and instead, Google merged Google Brain—its other AI team—with DeepMind.

“While DeepMind may have been unhappy to work on military AI or defense contracts in the past, I do think this isn’t really our decision anymore,” said a DeepMind worker to TIME.

Google has published its AI principles and assured that Project Nimbus is not “directed at highly sensitive, classified, or military workloads relevant to weapons or intelligence services.”

However, a worker from DeepMind told TIME that the company’s response doesn’t confirm if it can enable violence and other negative developments, and it “is so specifically unspecific that we are all none the wiser on what it actually means.”

Since the letter began circulating in May, Google hasn’t put any of the suggestions into force, four people with knowledge of the situation told TIME. “We have received no meaningful response from leadership, and we are growing increasingly frustrated,” one of them said.

Perplexity AI to Launch Ads on Search Platform by Year-End - 2

Photo by Glenn Carstens-Peters on Unsplash

Perplexity AI to Launch Ads on Search Platform by Year-End

  • Written by Andrea Miliani Former Tech News Expert
  • Fact-Checked by

Perplexity AI, the AI-powered conversational and research company, will introduce advertising on its platform by the fourth quarter.

According to Reuters , the Nvidia and Jeff Bezos-backed company launched a program for publishers in May to share revenues for interactions. The news agency highlighted Fortune, TIME, and Der Spiegel among its initial partners.

OpenAI has also recently announced partnerships with publishers like Conde Nast , News Corp, and TIME , but advertising agreements were not disclosed.

Perplexity AI is taking a step forward into acknowledging the publishers’ value and facing recent plagiarism allegations and controversies.

According to CNBC , multiple media outlets, including Forbes, have accused Perplexity AI of sharing plagiarized versions of its stories without referencing its websites or brand names.

Perplexity AI told CNBC that it has made changes to its model to ensure proper citation. During the debut of its “Publishers Program,” Perplexity AI shared data and statistics to encourage more publishers to join.

In its pitch deck, the AI company noted that Perplexity AI has over 2 million downloads, receives over 230 million queries every month, and has seen an eightfold increase in queries from the United States.

In a July interview with CNBC, Dmitry Shevelenko, Perplexity’s chief business officer, explained that if the AI considered three articles from the same publisher to provide one answer, the Publishers Program’s member would receive “triple the revenue share.”

Perplexity AI’s goal is to reach over 30 publishers by the end of the year. Entrepreneur, WordPress, and The Texas Tribune have also shown interest in the program. According to an anonymous source, the company will follow a cost-per-thousand-impressions (CPM) model of around $50, considerably higher than the traditional rates of $2.50 per CPM on desktop and $11.10 on mobile videos reported by Semrush last year.

OpenAI also just launched a new AI-powered search engine to compete against Google and other AI search engines.