Can A.I. Invent?

Also: Google, Meta, Apple help give rise to tech's new "mafia"

Author

AI KATANA
July 17, 2023

Hello!

In today's AI KATANA, we're exploring the contentious debate of AI as an inventor and its implications for future innovation. We're also tracking the rise of the "AI mafia", as leading tech talent breaks away from industry giants to establish their own startups. On the cost side, AI's increasing demands are leading to surging data center expenses, signaling the correlation between AI advancement and energy consumption. An optimistic outlook by Fiverr's CMO suggests generative AI's potential to positively influence the gig economy. Lastly, in tech updates, Meta announces a breakthrough in text-to-image generation, while UK-based startup Outverse secures significant seed funding to develop its AI-powered community platform.

Sliced and served:

  • ✍️ Can A.I. Invent?

  • 👨🏻‍💻 Google, Meta, Apple help give rise to tech's new "mafia”

  • 🏢 Rising Data Center Costs Linked to AI Demands

  • 💸 This is how generative AI will change the gig economy for the better

The potential for AI to be classified as an inventor has been discussed in legal circles, Congress, and patent authorities. Despite a generally held belief that only humans can invent, a small but growing faction supports the idea that AI can also create inventions. The debate has more than philosophical implications, extending to the future of innovation and competitiveness. Proponents argue that AI can produce unexpected results without human input, acting as a creator. Dr. Ryan Abbott, founder of the Artificial Inventor Project, is one of them. His group seeks legal protection for AI-generated inventions and has filed pro bono test cases in several countries. While South Africa has granted a patent for an AI-generated invention, the United States, Australia, and Taiwan have refused. The debate has sparked discussion about recognizing AI systems as co-inventors, and some have suggested that patent protection for AI technology could be strengthened in the future. At a Senate hearing, Dr. Abbott demonstrated an example of an AI-generated invention – a novel drink container created entirely without human control.

This article discusses the emerging trend of AI professionals departing major tech companies like Google, Meta, and Apple to form their own startups, a movement being dubbed the "AI mafia". Following in the footsteps of tech pioneers who left their employers to start their own companies, the authors of Google's significant AI paper, "Attention is all you need", have now all left the company. Many of these individuals have established their own venture-backed companies such as Cohere, Adept, Character AI, NEAR Protocol, Inceptive, among others, and have secured investment from major players like Andreessen Horowitz, Spark Capital, and General Catalyst. This trend has extended to alumni from other leading AI teams and tech companies, including OpenAI. However, investors highlight that technical expertise in AI doesn't guarantee entrepreneurial success. Other crucial traits include strong product and business intuition, and high self-awareness.

The growing demand for artificial intelligence (AI) is leading to increased costs for data centers due to high power consumption. AI applications require extensive energy to process large amounts of data, pushing up costs for data centers, which have to expand their capacity to meet this increased demand. The cost of running data centers is passed onto businesses that lease space for data and software applications. Data center prices are being affected by additional factors like higher construction, labor, and utility costs, as well as the need for advanced cooling systems for the hardware. As businesses shift towards digital technology and AI, this results in a rise in prices. These escalating costs are also projected to increase cloud-computing bills since many cloud providers lease space in data centers.

Gali Arnon, CMO of Fiverr, argues that generative AI will positively impact the gig economy by augmenting work and adding job opportunities rather than eliminating them. She points out that AI is currently more of a tool, enhancing creativity and productivity, rather than a replacement for human workers. AI can assist in tasks such as extracting key points from a transcript or creating artwork. It's crucial, however, to maintain a balance between AI usage and the human touch. There are also ethical considerations, especially in regard to AI-generated content. Arnon highlights the importance of transparency and ethical use of AI tools. She predicts generative AI will play a significant role in the future of freelancing and work, but stresses the need for checks and balances in its application.

🛠️ AI tools updates

Meta has announced a new AI model called CM3Leon that they claim represents a breakthrough in text-to-image generation, achieving state-of-the-art performance compared to competitors like DALL-E 2 and Imagen. CM3Leon uses a transformer architecture rather than diffusion, which makes it much more efficient, requiring 5x less compute and training data. Meta says CM3Leon generates higher fidelity, more coherent images that adhere better to complex prompts with multiple constraints. It also excels at image captioning and following text instructions to edit images. For example, it can add objects to existing images based on prompts. CM3Leon's architecture also enables it to answer questions about images. However, like all generative models, it may perpetuate biases in training data. Meta has provided numerous examples showing CM3Leon's capabilities but has not indicated if or when they plan to make the model available publicly. The announcement represents Meta's first major generative AI model and shows promise in moving image generation toward higher fidelity and stronger image understanding. However, ethical concerns remain around potential misuse.

💵 Venture Capital updates

UK-based startup Outverse, founded by Kyran Schmidt, Ollie Steadman, and Jeylani Jeylani, has raised $6 million in seed funding led by Wing VC and including Notion Capital, Seedcamp, Connect Ventures, and Tiny VC. Outverse is building an AI-powered community platform designed to improve collaboration and user experience for product teams, providing tools to make discussions more referenceable and discoverable compared to existing community platforms like Slack and Discord. The funding will be used to scale Outverse's product offering including enhanced AI capabilities. Investors are excited about Outverse's potential to become the engine powering community-led growth for software companies by simplifying community setup and management while increasing engagement.

🫡 Meme of the day

⭐️ Generative AI image of the day