Activities
Categories
Programs
Activities
Categories
Programs
FP Special: Amit Luthra, Managing Director(ISG), Lenovo, discusses the importance of autonomous data for independent AI
Nearly 93% of Indian corporations aim to put funds into AI in 2024, with their focus not solely on GenAI and LLMS, but also on AI inference models. In this context, the formation of independent AI is essential, but achieving this without autonomous data would be very challenging.
Artificial intelligence has indeed made remarkable progress, but what we've seen so far is just a small part of its full potential. Our attention has mostly been on generative AI and LLMs, however, significant advancements are occurring in various other areas at a quicker pace.
For instance, 93% of Indian corporations intend to invest in Artificial Intelligence by 2024, with their interest extending beyond General AI and LLMS to AI inferencing models. These models assist them in examining and interpreting scenarios and determining appropriate responses. While General AI has sparked both amazement and fear, the future for Indian businesses lies in analytical AI, automation, edge-computing, and AI inference.
We had a conversation with Amit Luthra, the head of Infrastructure Solutions Group at Lenovo India, during the launch of their 'CIO Playbook 2024 – It's all about Smarter AI' report. We discussed the potential future of AI in India.
We cover a variety of topics, from the concept of data sovereignty to the ethical considerations surrounding AI. We also debunk the myth of AI "taking" our jobs and explore how CIOs intend to utilize AI for more than just its trending status, but for its capacity to bring about significant and impactful changes. Below are the revised snippets from our discussion:
What changes have occurred in India's computational requirements over time, particularly concerning AI? India's digital shift was built on four main elements – Social Media, Mobility, Analytics, and Cloud Computing. AI was always a part of Analytics. Most corporate applications have attempted to incorporate Artificial Intelligence. However, two significant events caused a shift in the situation.
Initially, the arrival of IoT led to a huge influx of data from sensors. Businesses recognized that they couldn't afford to collect the data first and then analyze it retrospectively to achieve their objectives. Suppose a manufacturing plant aimed to enhance their production quality. They needed to scrutinize their data in real-time, particularly if the data was time-sensitive. This situation amplified the demand for Edge Computing.
The arrival of GenAI, along with the development of large LLMs, would be the next step. To construct these extensive LLMs, you require advanced systems with immense computational capabilities.
Nonetheless, I believe that although a substantial portion of the present investments in India are concentrated on developing LLMs, we will need to divert our attention to boosting investments in AI inferencing in the future. The rate of growth for inferencing will be significantly higher. It will also involve a combination of CPU-GPU workloads, instead of solely depending on GPU-based systems.
What are the primary obstacles faced by businesses, particularly startups, when implementing AI? Especially when dealing with generative AI, there needs to be an organized structure. The most significant hurdle is maintaining effective data management and ensuring the accuracy of data. If the AI models are given poor quality data, they will generate poor quality results. It's possible to develop an impressive LLM that learns rapidly and has excellent algorithms. However, if the training data is incorrect, incomplete, outdated, or of low quality, it will cause problems later on.
Creating an analytics structure that aligns with certain objectives is also vital. Businesses must understand the purpose of requiring an AI model and their intended results. They should question, "What is my goal with the collected data?"
Many AI initiatives at various companies don't succeed due to a lack of definite purpose for the AI application. The ultimate goal for certain firms might be to enhance customer happiness, while others might aim to boost manufacturing.
Some might aim to establish cleanliness routines to boost efficiency, while some may just aim to avoid robbery and examine scams. Companies must be clear about their ultimate goal and what they intend to accomplish. When this is determined, the appropriate technology can be easily implemented.
Subsequently, the aspects of safety and morality need to be equally emphasized. When working with information, it can either make or break you. Companies must question how they should tackle the moral element and where the boundaries should be set. Further, they need to make certain that they handle sensitive information correctly, maintaining appropriate levels of privacy and security.
The CIO Playbook 2024 indicates that nearly half of all CIOs face difficulties when trying to fill AI-related positions. Despite India boasting the highest number of engineers and developers, there seems to be a gap between industry expectations and the existing skill set of these professionals. AI has been a part of the engineering curriculum for over 25 years. However, translating academic knowledge into practical application in the modern corporate environment is a different ball game. It is unrealistic to assume that recent graduates can immediately demonstrate proficiency in AI.
The problem isn't a lack of resources, but rather creating a supportive structure for these resources.
This becomes particularly difficult as situations can shift rapidly. What was thought to be essential last year may be irrelevant this year. Presently, a business owner may wonder how they can provide ongoing guidance and training to their teams, and how they can consistently promote a learning environment.
The problem of retaining talent also exists. Employee poaching is a critical concern, particularly in the field of artificial intelligence. In this situation, companies must consistently allocate funds to their assets and possess the determination to maintain them.
Discussions surrounding Sovereign AI and Data Sovereignty are quite prevalent these days. Given the current progression of LLMs, what obstacles do you foresee in establishing sovereignty? Data sovereignty pertains to the awareness of your data's location and confirming that it is kept within your nation's geographical limits. Usually, data might be saved on the cloud, but the physical server of this cloud could be located in a different country.
For instance, numerous multinational corporations previously maintained centralized data centres situated anywhere globally to enhance and streamline operations. In such a scenario, you can't attain sovereignty as the data, potentially comprising personal details of Indian nationals, is stored beyond the nation's borders.
In this situation, the importance of data sovereignty legislation cannot be overstated. It mandates that personal information of Indian nationals, for instance, must remain within the country.
After confirming this, the security and analytics structure, along with the government, take over. These structures must make certain that crucial information is not exposed, and that it is managed appropriately and preserved according to the rules.
So, how do businesses in India utilize this information and build an analytics structure around it? It's crucial to understand your goal and the end result you're aiming for. Each sector and situation varies.
What types of LLMs or applications could potentially be created in this situation? Here's an example for clarification. If I need to contact the customer service team of a service I'm using, I'd rather call them directly instead of interacting with a chatbot on an app, mainly due to the generally unsatisfactory experience. However, we're now witnessing the rise of AI-powered voice assistants in customer service roles, where the entire process is voice-activated.
A good illustration would be the retail sector. In the near future, traditional sales representatives could be replaced by digital avatars. These avatars will be designed to handle your unique inquiries, tailored to your specific needs. Imagine you visit a car dealership and request a rundown of the features of a new vehicle. If the salesperson doesn't have much experience, it could lead to misunderstandings for the potential buyer. Furthermore, there's a possibility that they might not be able to respond to your questions if you ask something highly specific and unusual about the product.
Consider this situation, but with a digital representation that is nearly identical to a human and is knowledgeable about the product it is tasked with. After a single training session, it could respond to all your inquiries.
Furthermore, the company doesn't have to dedicate a significant amount of time to retrain and reorient employees whenever new updates occur. In this scenario, the only requirement is to confirm that the language models, regardless of their size, are updating in real-time and have access to the most recent data sets. Think about how dissimilar the previous versions of ChatGPT were compared to the GPT4.5-enhanced ChatGPT.
After the LLM is constructed, it's critical to ensure that it's easily transportable, current, and most importantly, simple to implement. This is where Lenovo steps in to assist companies with a variety of solutions that span from handheld devices to cloud-based systems, as AI is not limited to a single LLM. Rather, AI encompasses the whole ecosystem.
The Indian government has not yet implemented rules regarding AI in the country. What would you, as a business, want these rules to include? Additionally, how do you think the Digital Personal Data Protection Act of 2023 will affect the growth of AI in India? The establishment of a robust security and ethics structure will be critical. How can we protect individuals from deepfakes, which we've already encountered? The focus of discussions should be on security and ethics. Understanding the landscape will help entities creating LLMs understand their boundaries.
Senior executives and business leaders, regardless of their proficiency in AI or technology, are enthusiastic about incorporating AI into their operations. How can Chief Information Officers (CIOs) engage with them to guarantee the proper and appropriate application of AI? The desires of businesses and the requirements of technology often diverge, as highlighted by our survey for the CIO Playbook 2024. Top-level executives tend to prioritize generative AI due to its popularity as a buzzword.
In 2023, Generative AI was a non-existent concept, but presently, it's a high-priority topic and a popular term. Last year, the focus was on customer experience and satisfaction metrics. Yet, this year, it has become the second most critical aspect as businesses have realized that attracting a new customer can be costly, but keeping an existing one is more manageable.
Likewise, income and earnings expansion, traditionally viewed as the top priority for all companies, has dropped to third place. Sustainability, previously not a major concern, has risen to the fifth place on the priority list due to moral considerations.
Let's now focus on what the Chief Information Officer (CIO) or the tech department of a company aims to prioritize. CIOs continue to hold the view that automation should be the topmost priority. Giving top priority to automation can resolve numerous issues. Therefore, Artificial Intelligence (AI), particularly generative AI, comes into the picture only after automation has been implemented.
For AI to effectively handle high-performance tasks, it requires a specially designed infrastructure. Additionally, it's crucial that security and ethical considerations are central to your AI. As previously mentioned, these elements act as a structure for AI, which can't function autonomously and must serve a purpose. The application of AI is strategic, involving various components within a defined framework. This is a concept we have always supported, and we are pleased to find it confirmed in this survey when we consult with CIOs.
What aspects should the government prioritise when integrating AI into their administrative tasks? How can AI enhance the efficiency of the government? The government has a clear direction regarding AI. Since we elected them, it's important for us to align with their plan. In meetings with us – the industry leaders at different forums, ministers and relevant authorities gather diverse perspectives and insights on various topics. They also encourage us to provide our viewpoints and suggest optimal practices.
I believe that once a government establishes an AI strategy and introduces structures such as data sovereignty, compliance will be necessary for everyone. Further, our Prime Minister, Shri Narendra Modi, has also expressed support for 'AI for All'. In Lenovo, we have been advocating 'AI for All' for several years. We also share the belief that an 'AI for All' approach would significantly strengthen our vision.
Ultimately, the burning question that everyone is pondering is: How will AI transform the workplace and the concept of jobs as we currently understand them? AI is merely another technological advancement or set of instruments designed to enhance human performance. Take this into account – there were warriors in World War I, World War II, and all subsequent conflicts. During World War I, these warriors wielded .303 rifles, but today, they're equipped with LMGs and MMGs. Despite the shift in weaponry and the massive technological progress since then, the need for warriors remains.
AI is simply another instrument similar to those weapons, another facilitator that will allow them to perform more efficiently and handle sophisticated machinery.
Indeed, AI will transform the characteristics and possibly the caliber of occupations, but it won't eliminate or render them obsolete. There may be a decrease in the need for certain types of positions, but it will also pave the way for new opportunities.
I'm of the opinion that artificial intelligence will lead to an increase in salaries, as it will require individuals to acquire new and advanced skills. To avoid becoming irrelevant, people will need to focus on personal development. Consider the transformation in the IT industry. Two decades ago, during job interviews, I used to inquire if applicants were familiar with SAP sizing and exchange sizing. However, these questions are no longer relevant today.
If an individual claims to be highly skilled in PowerPoint on their resume but lacks knowledge on its integration with AI, their skill set is not up-to-date. AI is a crucial tool for our advancement. Ultimately, it's about maintaining relevance, securing employment, and aligning with industry needs. To achieve this, self-investment is key.
Look for us on YouTube
Top Programs
Related Articles
NVIDIA's Jensen Huang says AI hallucinations can be fixed, estimates full artificial intelligence will be here in about 5 years
Apple finally unveils MM1, its new AI system for generating text and images
Microsoft recruits Mustafa Suleyman, cofounder of DeepMind, to head up their new consumer-focused AI division
Samsung and Rebellions, chip makers from South Korea, aim to outdo NVIDIA
AI hallucinations can be remedied, says NVIDIA's Jensen Huang, predicting comprehensive artificial intelligence in roughly 5 years
After much anticipation, Apple rolls out MM1, its AI system designed for creating text and images
DeepMind cofounder Mustafa Suleyman has been hired by Microsoft to spearhead their new consumer AI department
South Korean chip producers, Samsung and Rebellions, are setting their sights on surpassing NVIDIA
Available on YouTube
Firstpost holds all rights, protected by copyright, as of 2024
+ There are no comments
Add yours