At the 26th Asian Investment Conference that was held in Hong Kong from March 21 to 23, 2023, Credit Suisse released a report on the potential of ChatGPT. Credit Suisse’s global sector research teams published a thematic research report that takes a deep dive into ChatGPT, adjacent AI use cases that may emerge by industry/sector, the AI hardware supply chain that will scale it, and the companies that are most likely to benefit on the back of sharply increasing AI model deployments.
The report highlights how the OpenAI software registered 100 million users in just two months, outpacing the likes of TikTok and Instagram in terms of growth rate. The AI chatbot became popular in just a matter of weeks — way faster than social media platforms, according to TIME. ChatGPT got one million users in only five days after it launched on November 30, 2022, 57 million users by December 2022, and 100 million users by January 2023.
The report stated that Generative Artificial Intelligence is likely to be transformative, after further fine-tuning, and eventually regulated. It said: “We view ChatGPT’s technology innovations, the progression of generative/conversational AI (ChatGPT is a product of generative AI), and Bing AI (among other apps/services) as broadly transformative and predominantly a productivity, cost-cutting, efficiency tool versus a revenue-generating tool today as it pertains to most industries.”
Also read: Is ChatGPT tasked to replace human jobs? This is what we know so far
ChatGPT’s ability to write and check code in various programming languages would help dramatically enhance the speed of innovation for software programmes, it said further, adding other industries that require professionals to search/validate facts or inquiries will also see a real-time benefit as ChatGPT is already a helpful tool for several productivity use cases, like an idea or content generation.
As per the report, ChatGPT’s LLMs are expected to be further fine-tuned over time with the next big milestone being GPT-4, a LLM with significantly more parameters versus GPT-3’s 175 billion.
However, despite ChatGPT (and other recent generative AI innovations) being embraced by businesses and organisations globally already, there are risks posted by ChatGPT and some limitations of the GPT LLMs have surfaced. “As such, we believe rules and regulations are needed for AI development and ChatGPT specifically considering the potential impact it has on society. Although there are no current regulations on ChatGPT yet, relevant discussions have been going on regarding how to make sure the impact from the recent developments are responsible and controlled,” it read.
Also read: What is ChatGPT-4 and how to use it: All you need to know
While it is given that the Information Technology industry is the key beneficiary of ChatGPT, the report found that within the technology industry, 30 percent of all new code is generated with AI assistance through tools such as ChatGPT and Copilot -- a testament to the value proposition of the technology and a material productivity accelerator.
Notably, the AI hardware and semiconductor supply chain is generally positioned as a beneficiary too. As the report highlights, “AI models are very compute intensive during training and continue to be compute intensive as users consume compute resources for inference (submitting a prompt into ChatGPT’s prompt box for example).
"For the next five years, there is likely going to be a significant scale-up effort to accommodate AI workloads and their required resources to deliver their development propositions.”
The report also discusses several sub-categories of the AI hardware and semiconductor supply chain that may not be on the typical investors’ radar.
The Asia/Europe technology supply chain for AI / ChatGPT will be a key facilitator in a rapidly growing AI world.
The Credit Suisse global tech team, in the report, summarised the supply chain implications and company-level beneficiaries from the rapid uptake of Chat GPT and its potential to further accelerate adoption for the AI ecosystem.
It said: “While the new ChatGPT workloads are not yet offsetting macro to drive upside in supply chain orders, we do view concentrated investments leveraged to the acceleration of AI having the ability to show over-indexed growth through the industry slowdown. In the medium-term, the uptake of AI services and its industry use cases for revenue generation and cost/capex efficiencies can feed to a new cycle of hardware and semiconductors to maintain innovation and advances.”
The CS report also highlighted how AI compute and memory can benefit the semiconductor sector. It pointed out that AI training and inference are compute-intensive tasks that should continue to drive semiconductor advances for compute, storage and the transmission of data. The data centre compute TAM including accelerators has maintained a 14 percent CAGR from 2019-24E, far outpacing the CPU server growth at a 2 percent CAGR.
Further, in semiconductors, AI has the potential to improve prospects for server memory for the memory leaders, now crossing over mobile at 40 percent of industry bits, power management into AI boards, network switch ICs and ASICs and IC design services.
Moreover, the supply chain beneficiaries from advances in compute intensity will be a good driver for leading-edge silicon, which is now replacing mobile as a key driver for innovation both on advanced manufacturing and high-end packaging integration, in our view.
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
