Really, Elon? Another Artificial Intelligence Venture?

The Gist

AI challenge: xAI's launch signifies Elon Musk's direct rivalry with AI giants, specifically Microsoft and Google.
AI skills gap: Salesforce's study reveals a significant generative AI skills deficit, underlines need for immediate upskilling in all sectors.

In his quest to create a formidable competitor to OpenAI, the AI firm that brought the popular chatbot, ChatGPT, into the limelight, and adding to his diverse tech portfolio, Elon Musk has announced the establishment of xAI, a new artificial intelligence venture.

According to xAi website, industry leaders from Google, Microsoft, OpenAI and Tesla are also part of the project, which will operate independently of X Corp (Twitter's parent company) with the goal to “understand the true nature of the universe.”

In a post on Twitter Wednesday, Musk said that he was “Announcing formation of @xAI to understand reality.”

Musk, who co-founded OpenAI and later left due to disagreements, has voiced concerns over the potential overpowering influence of AI. In March he was among a group of technology and business leaders — including Steve Wozniak and tech leaders from Meta, Google and Microsoft — to sign a letter that called “on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.”

However, the following month Musk, the CEO of Tesla, Twitter and SpaceX, announced his intention to start a new AI project called "TruthGPT" or "Maximum Truth-Seeking AI," with the goal of incorporating advanced AI algorithms that would sort fact from fiction to combat the spread of misinformation and fake news. And during a WSJ CEO Council event in May, he expressed the need for a "significant third horse" in the AI race.

The launch of xAI is seen as Musk's direct challenge to other leading AI giants, namely Microsoft and Google. Microsoft earlier this year deepened its ties with OpenAI through a substantial multi-year investment, whereas Google has been keeping pace with the introduction of its chatbot, Bard.

Have questions? The xAI team will host a Twitter Spaces Chat to discuss the new venture tomorrow.

In other AI news...

Salesforce Uncovers the Hidden Iceberg of Generative AI Skills Gap

Salesforce has released a study highlighting the significant generative AI skills deficit in today's business landscape, calling for immediate upskilling across all sectors. Their international data suggests that while two-thirds of business leaders are actively exploring the use of generative AI, a staggering 65% of senior IT officials confess that their companies are not equipped to deploy this technology. The primary hurdle is a widespread skills gap, with recent studies indicating that over 60% of the workforce feels inadequately trained to safely and proficiently utilize generative AI.

The research reveals that the current digital skills gap leaves less than half of employees confident in leveraging AI's full potential — and when it comes to generative AI, 67% of workers expect employers to help them learn how to use the technology, but two-thirds say that isn’t happening today.

In tandem with the study, Salesforce also announced Trailhead, Salesforce’s free online learning platform, aimed at everyone from “seasoned tech veterans to those just getting started in the industry to learn AI skills at their own pace.”

“Employees who can harness the power of generative AI can get more done faster and spend extra time on rewarding, meaningful work,” Ann Weeby, SVP of Trailhead, Salesforce, said in a statement. “They’ll be at a major advantage when it comes to serving customers, driving innovation, and making their companies stronger.”

Baichuan Bets Big on AI, but China's New Rules Could Hold Back the High-Roller

Chinese start-up Baichuan Intelligence, founded by Sogou creator Wang Xiaochuan, has reportedly unveiled its next-gen large language model, Baichuan-13B. Billed as one of China's leading LLM developers, Baichuan raised $50 million from angel investors within months of launching. The 13 billion-parameter model, trained on Chinese and English data, is open source and optimized for commercial use. In terms of scale, Baichuan-13B is comparable to Meta's LLaMa, with its training based on 1.4 trillion tokens.

However, Baichuan's rapid progress comes as China prepares for stringent AI regulations, potentially impacting the nation's competitive standing in the AI industry. Purportedly, China's forthcoming regulations on generative AI will focus on content, implying a more rigorous control than rules implemented earlier this year. This could mean businesses will need licensing before launching large language models, potentially slowing China's AI development.

Related Article: Google DeepMind CEO Says Company's AI Will Surpass ChatGPT

AI Tweet of the Week: Heidi Trailer: AI's Latest 'Spaghetti Moment' Sends Shivers Down Swiss Spines

Welp...while AI generated video seems to have improved since Will Smith first ate spaghetti, it’s still a bit...off. Swiss comedian and director Patrick Karpiczenko (@Karpi) tweeted that he “can never sleep again” after viewing the result of his request for AI to generate a trailer for a “Heidi movie.” I have few words for this...except maybe...yikes.

I've asked an AI to generate a trailer for a HEIDI movie and now I can never sleep again pic.twitter.com/8M9t726hrI
— Karpi (@karpi) July 10, 2023

AI Video of the Week: ITU Lets Robots Do the Talking at AI for Good Global Summit 2023

In a twist on the typical press conference, the International Telecommunication Union (ITU) said theirs would be “quite different.” During the AI For Good Global Summit 2023, ITU noted that, instead of multiple spokespeople from the ITU, they invited nine robots to the table, in an effort to showcase the capabilities of state-of-the-art robotics and how they could support the United Nations Sustainable Development Goals (SDGs).

Learning Opportunities

Webinar