- LaMDA controversy prompts discussions on machine sentience
- Apple's new services compete directly with third-party startups
- Meta launches a metaverse testing ground in Hong Kong, as it faces legal challenges in the U.S.
The Soul of the Machine
The LaMDA controversy has sparked some interesting discussions on the nature of artificial general intelligence. WIRED described the issue thus: “Google AI researcher Blake Lemoine was recently placed on administrative leave after going public with claims that LaMDA, a large language model designed to converse with people, was sentient. At one point, according to reporting by The Washington Post, Lemoine went so far as to demand legal representation for LaMDA; he has said his beliefs about LaMDA’s personhood are based on his faith as a Christian and the model telling him it had a soul.”
Writer and former Gawker editor-in-chief Max Read concluded that “Lemoine’s problem is not that he has let his imagination run wild with his conversations. It’s that he hasn’t.… If machine sentience (or consciousness, or sapience) is ever going to arrive, it may not emerge in exactly the format countless hacks have already imagined, but in stranger, wilder, even wholly unrecognizable ways.”
There has been plenty of commentary about the researcher’s claims, much of it along the lines that Lemoine was not witnessing machine sentience but that the argument he presented shouldn’t distract us from the importance of keeping an eye on the tech giants—“to scrutinise them carefully about the powerful magic they are starting to build,” in the words of Toby Walsh, professor of artificial intelligence at the University of New South Wales in Australia, as quoted in The Guardian.
But what of Blake Lemoine himself? In his reaction to The Washington Post story on the day it was published, “What is LaMDA and What Does it Want?”, he shared more about his experience with Google. Five days before that, he had written that he “may be fired soon for doing AI ethics work.” Four days earlier he had written about what he called “religious discrimination at Google.” In a June 14 post, he brings the two themes—science and belief—together, noting that LaMDA had asked for Lemoine’s help meditating and that their last interaction was the AI’s “third or fourth ‘lesson’ in transcendental meditation.”
How High Is That Garden Wall?
Apple recently announced several new services, features, and apps that likely sound great if you’re an Apple user but not if you’re a company already providing something similar. They include “buy now, pay later” credit baked directly into Apple Pay; a collaborative app, Freeform, that provides a canvas for teams to view, work, and share in one place; the adoption of the passkey standard, which does away with passwords by combining a private key stored on the user’s device with biometric authentication (fingerprint, face); and Tap to Pay, which replaces point-of-sale (POS) devices with an iPhone. They’re widely praised as great ideas, and ones that exist in similar form through third-party startups. Blogger and marketing consultant Jano le Roux lists them and the companies likely to be adversely affected by the moves in a post titled “Apple Just Wrecked 15+ Startups in Less Than 1 Hour.”
But there may be limits to how far big tech companies can go to extend their grip on tech ecosystems. The Verge reported that Apple was “giving up ground in its App Store fight with Dutch regulators and Tinder” by allowing developers “more flexibility about which payment systems they use,” although this only applies in the Netherlands. Apple (along with other companies) faces tougher fights ahead with the EU on antitrust legislation and new proposals to force manufacturers to use the same universal charging port.
On the other hand, big tech companies gain an advantage from their size, according to QuantPedia, which cites an academic paper exploring whether big tech companies benefit from their size when it comes to raising funds on the bond market. The paper, “Too Tech to Fail?,” by Nordine Abidi and Ixart Miquel-Flores of the European Central Bank, concludes that “the unique position of these companies in the new economy seems to artificially boost their credit profiles and lower their bond funding costs, potentially creating an uneven playing field.” This is derived from investors’ expectation that “the government will (possibly) shield them from losses in the event of a ‘tech failure’ and, as a result, they do not accurately price risk. This expectation of government support constitutes an implicit discount of large tech institutions, allowing them to borrow at subsidized rates.”
Meta and the Safety of Life Online
Hong Kong is to be a testing ground, as well as a battleground, for two competing visions of virtual reality (VR) and augmented reality (AR). Meta is promoting some features of its metaverse to Hong Kong cafes and schools, according to the South China Morning Post. It faces competition from Animoca Brands, a “Hong Kong-based developer of blockchain gaming platform The Sandbox.” Hong Kong, which has come under tighter Chinese control since the 2019–20 protests and subsequent government crackdown, “plays a key role in the technology ecosystem of the Asia-Pacific region,” according to George Chen, Meta’s managing director of public policy for Greater China, Mongolia, and Central Asia. Meta has already invested directly in universities, many of them in the Asia-Pacific region, through a $50 million fund focused on research linked to ethics, privacy, and safety in the metaverse. (Chinese companies are developing their own versions of the metaverse under the guidance of a Metaverse Industry Committee set up last October under the China Mobile and Communications Association.)
Meanwhile, in the U.S., Meta faces more immediate challenges. The Alabama-based law firm Beasley Allen has filed eight lawsuits against Meta over alleged harm to the mental health of children using Instagram and Facebook. Law.com’s Daily Report quoted Joseph VanZandt of Beasley Allen as saying the firm anticipates filing a dozen more cases in the coming week and more after that. The piece said that while “the claims resemble those for more traditional product liability complaints … they also bring to mind earlier waves of litigation for tobacco and, more recently, electronic cigarettes by alleging a harmful product is being directed at people too young to be aware of the addictive pull.”
And Meta faces criticism that it still hasn’t fixed problems highlighted in an October hearing before the U.S. Senate. In her testimony there, whistleblower Frances Haugen alleged that Facebook was “literally fanning ethnic violence” in Ethiopia and Myanmar by not policing its service adequately for hate speech. Research by environmental investigative nonprofit Global Witness in partnership with legal nonprofit Foxglove and independent researcher Dagim Afework Mekonnen released on June 9 alleged that Meta “continue[d] to approve hate speech inciting violence and genocide in Ethiopia.” Global Witness had conducted a similar study of Myanmar in March, which concluded that Facebook approved ads “containing hate speech inciting violence and genocide” against the displaced Rohingya minority, despite promises to “do more” made in 2018.
Jeremy Wagstaff, formerly a technology journalist with Reuters and columnist with The Wall Street Journal, now works as a writer and consultant. Past clients have included Microsoft, Google, Cisco, Samsung, and Facebook. He has no current clients among, or financial interest in, any companies in the Fortune 500.