5 things about AI you may have missed today: Son recreates late father’s voice via AI, Yamaha AI piano, deepfakes, more
Even as we enter the weekend today, December 23, interesting tidbits from the world of artificial intelligence keep coming in. In a shocking development, Infosys has lost a deal worth $1.5 billion with an undisclosed AI firm. The announcement came just days after Infosys Chief Financial Officer (CFO) Nilanjan Roy’s resignation. In other news, a group of teachers worked with Yamaha to create an AI-assisted piano that tracks the notes of the music and augments the performance by adding whatever keys are needed but not pressed. And, Arizona creates its own deepfakes to prepare for polls. This and more in today’s AI roundup. Let us take a closer look.
Infosys loses mega deal with AI firm
Infosys recently terminated its Memorandum of Understanding (MoU) with an undisclosed global AI company, just two weeks after the unexpected departure of CFO Nilanjan Roy, reported Hindustan Times. The deal, aimed at developing artificial intelligence solutions to address industry needs, was initially set for a 15-year commitment and was officially established in September 2023.
Infosys said in its exchange filings, ““This is in continuation to the disclosure made by Infosys vide letter dated September 14, 2023, titled “Company update” with respect to a Memorandum of Understanding with a global company which was subject to parties entering into a Master Agreement…The global company has now elected to terminate the Memorandum of Understanding and the parties will not be pursuing the Master Agreement”.
Yamaha develops AI-assisted piano
Musicians with disabilities, including Usami, who has cerebral palsy, used an AI-powered piano called “Anybody’s Piano” during a performance of Symphony No. 9 with the Yokohama Sinfonietta at Suntory Hall, as per a report by AFP. The piano, developed in collaboration with Yamaha, tracks the notes of the music and complements the performance by adding keys as needed. Usami’s dedication to playing with one finger inspired the AI program. The concert, featuring the innovative piano, marked a significant achievement for inclusivity in music.
OpenAI enters discussions for funding at $100 Bn valuation
OpenAI is reportedly in early discussions to secure a new round of funding at a valuation of $100 billion or more, Bloomberg reported. The details, including the terms and timing of the funding, are still in preliminary stages and subject to change. Additionally, OpenAI is said to be in talks with G42, based in Abu Dhabi, about funding for a new chip venture. The report does not clarify the relationship between the chip venture funding and the broader funding for the company, but it mentions discussions of raising $8 billion to $10 billion from G42.
Son recreates late father’s voice using AI
A St. Louis man, Phillip Willett, used AI to recreate his late father’s voice as a Christmas gift for his mother, as per a report by ABC News. As a content creator for The Content Guy media company, Willett had the idea suggested by his wife, who knew about his use of AI in his professional work. After discovering a community online that successfully used AI voice software for similar purposes, Willett decided to give it a try. Using ElevenLabs’ text-to-speech software, he recreated his late father’s voice and compiled a video book with photos of his father and his mother, which has gained viral attention on TikTok with over 3 million views.
US state creates its own deepfakes hoaxes to prepare for elections
Arizona’s Secretary of State, Adrian Fontes, is taking a proactive approach to address the potential misuse of artificial intelligence in election disinformation, according to a report by Politico. Following concerns about election fraud conspiracy theories in the 2020 presidential election, Fontes is leading a series of exercises to prepare for potential threats in the 2024 elections. In a recent two-day simulation involving around 200 stakeholders, Fontes used open access AI tools to generate audio and video deepfake content, attempting to deceive participants with misinformation. The goal is to assess and prepare for the risks associated with AI technology in spreading election-related disinformation.
