The Rise of Vertical AI

Also: UN council to hold first meeting on potential threats of AI to global peace

Author

AI KATANA
July 07, 2023

Hi!

Today's digest sees a continued focus on the intersection of AI and various sectors, with developments in Vertical AI, medical applications, and transformative shifts in software development roles, among other things. Starting off with the rise of Vertical AI, we delve into an industry-specific approach to artificial intelligence. This is then followed by a first-ever meeting by the U.N. on AI's potential threats to global peace, shining a spotlight on the necessary considerations around autonomous weapons and AI's role in nuclear weapon control. We also see AI's potential in transforming medicine, not as a replacement but as a supportive tool that can democratize information and enhance medical practices. Moreover, the advent of Generative Engineering illustrates the changes AI is bringing to the developer role, indicating a move into a post-training era of development. In other updates, OpenAI has announced general availability for its GPT-4 API, and older models will be phased out. And finally, from the venture capital world, Neko Health, co-founded by Spotify founder Daniel Ek, has secured $65M to expand its AI-driven preventive healthcare. Stay tuned for more insights and developments in the world of AI.

Slice it up:

  • ↕️ The Rise of Vertical AI

  • 🇺🇳 UN council to hold first meeting on potential threats of AI to global peace

  • 🏥How A.I. and ChatGPT May Change Medicine

  • 👨🏻‍💻 The rise of GenEng: How AI changes the developer role

The article elaborates on the emergence of Vertical AI, a progression from the well-established vertical SaaS (vSaaS) trend, where cloud-based software is customized for specific industries. As AI evolves, the prediction is that AI will be utilized in a vertical-focused manner, allowing for AI platforms that are developed based on industry-specific datasets. The article identifies two potential opportunities for Vertical AI: the creation of new AI-native vertical applications and the incorporation of AI features into existing vertical SaaS leaders' platforms. The winning entities in this vertical AI progression will likely be those who can effectively access industry-specific data, train large language models using these datasets, and deliver significant utility to customers swiftly.

The U.N. Security Council, led by the UK, is set to host its inaugural meeting on the potential threats artificial intelligence (AI) might pose to global peace and security. This initiative comes amid increasing concerns around the use of AI in autonomous weapons and nuclear weapon control. Secretary-General Antonio Guterres and international AI experts are expected to participate, providing insights into the profound implications of AI. The Council aims to establish a multilateral strategy to manage AI's opportunities and risks. While the technology can enhance development programs, humanitarian aid, peacekeeping, and conflict prevention, it also raises serious security questions. This meeting complements wider international efforts, like the EU's recently approved comprehensive AI regulations and OpenAI's call for a regulatory body overseeing AI systems.

During their intern year, the author would resort to a medical reference book when stumped by a tough question, assuming a competent clinician wouldn't need to look up answers. Now, however, they find themselves amidst the dawn of artificial intelligence (AI) in medicine, capable of writing patient notes, diagnosing, and communicating. Despite doctors' general hesitation to adopt new technology, the potential of AI, especially with advanced natural language processing, is becoming too compelling to ignore. AI's ability to recognize complex patterns, serve as an electronic scribe, and synthesize data much like an expert has potential to enhance the doctor-patient relationship and improve overall medical practice. Examples of AI's prowess include its ability to discern more from an ECG than a human expert and its potential for rapid diagnosis based on disparate symptoms. However, integrating AI into the medical field raises questions about the testing of such technologies and the possible erosion of traditional diagnostic skills in doctors. Despite these concerns, the author believes that AI can serve as an invaluable tool that can democratize information and improve medical practice, without replacing the essential human aspects of being a doctor.

The rise of Generative Engineering (GenEng) signifies a fundamental shift in the role of developers, moving us into a "post-training era." Here, developers will not necessarily need to be experts in deep learning, but rather, they will focus on integrating large language models (LLMs) into enterprise application architectures. This democratization of AI is set to expand the field of AI innovation to the broader community of developers. The GenEng practitioner, equipped with traditional development skills and a deep understanding of how to best leverage and integrate AI technologies into applications, will be the driving force behind this revolution. The future of software engineering will no longer be solely deterministic, but will include probabilistic solutions generated by AI, altering problem-solving approaches, team structures, and overall business functionality.

⚒️AI tools updates

OpenAI has announced general availability for its GPT-4 API and the retirement plan for older models in the Completions API by the beginning of 2024. Also, the GPT-3.5 Turbo, DALL·E, and Whisper APIs are now generally available. With the aim of improving support for diverse use cases, developers are encouraged to switch to the Chat Completions API, which now accounts for 97% of GPT usage. The Chat Completions API offers higher flexibility, specificity, and a more robust defense against prompt injection attacks. Older completion models will be phased out, with base GPT-3 models being automatically upgraded to new versions. Developers using other older completion models will need to manually upgrade their integration by specifying GPT-3.5 Turbo in their API requests. Moreover, fine-tuning capabilities are expected to be enabled for GPT-4 and GPT-3.5 Turbo later this year, with priority given to those who have fine-tuned older models. Lastly, users of older embeddings models and the Edits API are also required to migrate to newer versions by January 4, 2024.

💵Venture Capital updates

Neko Health, an AI-driven healthcare technology company co-founded by Spotify founder Daniel Ek and Hjalmar Nilsonne, has raised $65 million in a Series A funding round led by Lakestar, with participation from Atomico and General Catalyst. The funds will bolster strategic expansion, research and development, clinical studies, and talent acquisition. Neko Health utilizes innovative medical scanning technology, including a 360-degree full-body 3D scanner, to gather comprehensive health data. The AI-powered system analyses the data, enabling early detection of health issues. The technology offers rapid, accurate, non-invasive preventive healthcare solutions. Having already completed over 1,000 scans at its inaugural Stockholm clinic, the company anticipates strong growth, underscored by 80% of customers prepaying for follow-up scans within a year.

🫡Meme of the day

⭐Generative AI image of the day