5 important things I learned from Generative AI landscape report 2023 from McKinsey by Gaurav Aug, 2023
Generative AI landscape: Potential future trends
Lizzi C. Lee is an Honorary Junior Fellow on the Chinese Economy at the Asia Society Policy Institute’s (ASPI) Center for China Analysis (CCA). She graduated from MIT’s Ph.D. program in Economics before joining the New York-based independent Chinese media outlet Wall St TV. She is currently the host of “The Signal Live with Lizzi Lee” powered by The China Project, where she interviews the most knowledgeable minds on China for analysis of the ever-evolving business and technology ecosystem. Qiheng Chen is a Senior Analyst at Compass Lexecon, where he provides competition economic analyses for mergers and litigations, particularly those involving the semiconductor industry and tech platforms. He has researched China’s laws and policies on tech regulation, data governance, and cybersecurity, and consulted on these topics. Qiheng is also a Young Economist Representative at the ABA Antitrust Section’s International Comments and Policy Committee, and an Honorary Junior Fellow on Technology and Economy at the Asia Society Policy Institute’s Center for China Analysis.
OpenAI conducts innovative research in various fields of AI, such as deep learning, natural language processing, computer vision, and robotics, and develops AI technologies and products intended to solve real-world problems. Some popular applications include image generation, text generation, medical image synthesis, drug discovery, content creation, language translation, virtual avatars in gaming and virtual reality, and fashion design. Additionally, generative AI is transforming customer service with intelligent chatbots and enhancing marketing strategies with automated content creation. In the generative AI application landscape, several prominent use cases stand out.
Infrastructure: Cloud Platforms –cloud deployment model and how it runs model training and inference workloads
It wasn’t just a joke that the article was co-written with GPT-3; it actually was. And then I’d be like, “Specifically for image generation, you can think of it as ….” That human-machine iteration loop I hadn’t experienced before, and it was very much how we created both the blog post and landscape. Then the other big category where there has been a lot has been in the text space. So there’s a lot of these marketing Gen AI companies, and some of them are really working.
Starting from random noise, Stable Diffusion models gradually transform it into meaningful data, such as an image or a piece of text. Despite their computational intensity, recent improvements have made these models increasingly accessible and applicable across various domains. Unique to Stable Diffusion models is their ability to generate samples at any point during the diffusion process, offering a blend of abstract and realistic outputs.
From Simple to Sophisticated: 4 Levels of LLM Customization With Dataiku
It can identify keywords and phrases for the target audience and include them in the content. You can use generative AI tools to improve the overall flow of content and rise in search engine rankings. Notably, other forms of generative AI actually create videos, images and other rich media content. The early reviews of initial efforts in this area reveal much work still needs to happen, but I think entrepreneurs need to be aware of the significant potential. Additionally, many make the argument that ChatGPT still requires more work to improve its overall accuracy.
- In a world where AI is no longer a distant concept but an integral part of our lives, understanding the nuances of generative AI models has become essential.
- Generative artificial intelligence (GAI) has taken the world by storm, with new adaptive tools revolutionizing how we work, learn, and interact with information.
- For example, a customer service bot could use generative AI to generate responses to customer inquiries, while a social media bot could use it to create posts or tweets.
- As generative AI continues to evolve, advancements in these areas will contribute to safer, more reliable, and ethically responsible AI systems.
- Second, the growing demand for personalized and unique content, such as in the fields of art, marketing, and entertainment, has increased the need for Gen-AI platforms.
The scarcity of high-quality Chinese language training data creates a bottleneck in the development of high-performing language models. The greater use of video-based marketing also means that the use cases for text-based content are narrower. End-to-end apps using proprietary generative AI models present numerous benefits. They are easy to use, providing user-friendly interfaces for content generation. They are often affordable or even free to use, scalable to accommodate many users and incorporate strong security measures for user data protection. These applications may exhibit bias, depending on the data they were trained on, and there could be privacy concerns as these apps may collect and use user data in ways unknown to users.
The AI Platform Strategy
Founder of the DevEducation project
A prolific businessman and investor, and the founder of several large companies in Israel, the USA and the UAE, Yakov’s corporation comprises over 2,000 employees all over the world. He graduated from the University of Oxford in the UK and Technion in Israel, before moving on to study complex systems science at NECSI in the USA. Yakov has a Masters in Software Development.
While generative AI tools like ChatGPT offer many benefits, there are also drawbacks that startup leaders should be aware of. ChatGPT has been known to produce inaccurate information or generate information that doesn’t match the user’s query. Due to the way generative AI models are trained, there is also an inherent risk of bias. While silos and prompt engineering can overcome some of these limitations, generative AI isn’t ready for applications that may involve sensitive customer interactions where small mistakes can create large issues. What we now call generative AI wouldn’t exist without the brilliant research and engineering work done at places like Google, OpenAI, and Stability. Through novel model architectures and heroic efforts to scale training pipelines, we all benefit from the mind-blowing capabilities of current large language models (LLMs) and image-generation models.
Dataiku’s vision was always to provide the platform that would allow organizations to quickly integrate new innovations from the fields of machine learning and AI into their enterprise technology stack and their business processes. The arrival of modern Generative AI and LLMs is perfectly in line with that original vision. Generative AI is a subset of artificial intelligence that focuses on creating and generating new content, such as text, images, and audio, based on input data. To help you take advantage of generative AI, Wizeline has created an overview of all the GAI tools currently available. Our Map of Yakov Livshits resource helps you identify strategic options, explore potential applications, and make informed decisions to transform your methodologies, products, and services into AI-native ones.
The shift to foundational models and few-shot learning will be interesting to observe, as it could impact the importance of large, fine-tuned datasets that previous business models relied on. We are excited for healthcare specific data tooling that will help companies leverage these new technologies. However, despite the massive opportunity, healthcare is slow to adopt new technology.
As a “hot” category of software, public MAD companies were particularly impacted. The silver lining for MAD startups is that spending on data, ML and AI still remains high on the CIO’s priority list. This McKinsey study from December 2022 indicates that 63% percent of respondents say they expect Yakov Livshits their organizations’ investment in AI to increase over the next three years. As an example, scandal emerged at DataRobot after it was revealed that five executives were allowed to sell $32M in stock as secondaries, forcing the CEO to resign (the company was also sued for discrimination).
We might see Chinese AI companies creating a business model that is essentially “service-on-the-front, AI software-on-the-back.” These companies will likely behave quite differently than traditional SaaS companies. While the Israel-based lab AI21 and the Canadian startup Cohere are also building large-scale models, China is the only actor aside from the U.S. and UK to have multiple labs building and releasing these models. China has also built its own AI frameworks, including Huawei’s Mindspore and Baidu’s PaddlePaddle. These frameworks are not compatible with dominant Western frameworks, such as PyTorch and TensorFlow, but there are conversion tools like Ivy that might bridge between these frameworks. But little known in the West, China is building its own parallel universe of generative AI.
End-user-facing generative AI applications interact with the end user, using generative AI models to create new content (text, images, audio) or solutions based on user input. These apps without proprietary models use open-source, publicly available AI models without developing or owning the models. Revenue cycle operations represent companies that help healthcare providers improve Yakov Livshits the amount they receive from insurance companies after submitting a claim. This is the largest space by far in the Admin category and a market with many players. Long-time players like AKASA and Olive AI exist and newer technology offerings, such as Adonis (emphasizing better UI/UX and billing OS) and Candid Health (developing an API-like billing solution), are emerging.
The accessibility of these resources also poses challenges, potentially leaving smaller players at a disadvantage compared to multinational corporations. Efforts to make LLMs more accessible and energy-efficient are ongoing but untested. For companies that have been forced to go DIY, building these platforms themselves does not always require forging parts from raw materials.