Blog

The other day I was reading about the recent trend of IIT Jee toppers opting for B Tech in mathematics or data science stream rather than computer science which they were opting for the past many years. It is a clear indication of how manpower in different fields is being valued in industry. These days we place premium on STEM education — science, technology, engineering and mathematics. Much before this, India has attached importance to science and mathematics. The evidence was the cut off marks at H Sc. level for different streams — the science stream demanding very high cut offs. STEM option creates a vibrant knowledge economy, and an individual is armed with market-dynamic opportunities.

However, there is flaw in thinking here. Liberal arts such as sociology, economics, sociology etc. are necessary to build up our holistic personality. The liberal arts invest us with critical faculties. They make us adapt to the environment. They give us the abilities to do heuristic thinking. Such individuals can take decisions without following mechanical algorithmic thinking.

Liberal arts give individuals insights which they cannot get just by a cost-benefit analysis. Cardinal Newman wrote two books called Idea a University (1852). They were Mahatma Gandhi’s favourite books . Newman advocated liberal education till a certain level — say the intermediate or SSC level before starting skill education in crafts and sciences. A university must have a college of liberal arts and sciences and professional schools. It makes the students broadminded. Liberal Arts give the kick to infer intent behind decisions, unobvious patterns and less evident issues.

The ‘liberalist’ is more malleable and is liberal in all senses. He has an operating system that values freedom of choice, multiple pathways emotional considerations and intellectual empathy. He has intuitive knowledge. In a post-AI world, he will go much farther than the formulaic.

New education policy thus allows science stream students to study humanities courses and vice versa. Birla Institute of Technology and Science (BITS) followed this method of education since long. Engineering students must have exposure to economics, and medical students must have exposure to psychology. In fact, education should be cross-disciplinary.

In civil services, the IAS cadre is a generalist or liberalist. The other civil services are IPS, IFS, IRS and so on. Here there is a mix of generalists and specialists. In industry also we have technocrats and generalists. In decision making process, at the lower level, more technical skill are necessary, but as you move higher in hierarchy, you require more conceptual skills. These days management courses have students from diverse backgrounds (degrees in different disciplines) but are given broad-based generalist training in their MBA degrees. Many good executives are, therefore, B-Tech, MBAs. However, over the years, there was overwhelming number of engineers in the MBA courses of IIM. Therefore, as an equalisation measure, the non-engineering degrees were given some added weightage.

Alphabet’s Google is adding AI tools from Facebook and Anthropic to its cloud platform. Google’s clients will be able to access Facebook’s Llama2 LLM and Anthropic’s Claude2 chatbot. Google’s customers will have freedom to choose an AI model that best meets their needs. The model could be Google’s own or from one of its partners. Google clients could have access to more than 100 powerful AI models. As we already know, Duet AI is a product in its Workspace Productivity Suite. It will be accessible to the public by the end of 2023. Users can avail of a Gen AI helper which responds to prompts to help create content on apps such as Google Docs, Sheets and Slides. Duet AI can take notes during videocalls and can send meeting summaries. It can translate captions in 18 languages. There is ‘attend for me’ feature which sends the tool to join meetings on their behalf. It can deliver messages and create a recap of the event. Google collaborates with GE Appliances and Fox Sports. That enables customers to create customised recipes and see a playback of a sports event.

Google has shown its readiness to work with other firms in AI area. Google at the same time promotes its own AI products and services as the finest options.

In Cloud Computing, of course, Google trails behind Amazon and Microsoft. Therefore, when its cloud offerings are infused with AI, Google will benefit from the widest array of products it is offering to its customers.

Google is planning to make its AI-powered suite available to corporate G-mail accounts at an additional cost of $30 a month per person. Google has already acquired GM and Estee Lauder as its cloud software customers.

Google has introduced custom-built AI-chips. It has unveiled an enterprise scale to watermark and identify images generated with AI-plus tools. The tool is called SynthID.

Microsoft caught the company off guard, and therefore, Google is rolling out its AI plans.

Google introduces products for enterprises that are infused with AI. One such offering is Duet AI that has recently been released ( in the last week of August 2023).

Duet AI in fact combines two models — generative language and generative imagery. As we know, generative language models learn the structures, patterns of the language through training data and by being exposed to different formats and contexts. Once a prompt is given, there is an inference.

Generative image models, as we know, produce new images. They use diffusion technique — a model learns by deconstructing a large number of images into noise. The noise is then mapped back to images. Once a prompt is given, the model resolves noise into a correlated image. It thus creates text-to-image content.

In short, Duet AI is generative AI that creates text and images. It facilitates office work. It adds to productivity of the business. Google Workspace is a suit of cloud computing. Duet AI is a part of Workspace.

Duet AI has been sufficiently tested. It can be used for client outreach, emails, job postings, project plans and summarization of content. It can create text-to-image content on Google Slides. It generates customer outreach email content in Docs, drawing from sheets carrying customer information. Even project plans can be made using sheets. Duet AI is a multi-tasker.

Duet AI has HELP ME feature that allows custom template in sheets by describing what one is planning to achieve. It thus helps in organising work. AppSheet has apps to simplify work.(with no coding). Thus it facilitates automation of workstreams.

Content on Workspace belongs to an enterprise alone. Such data is never shared by Google. An enterprise is at liberty to delete its content or export it. The data cannot be used as raw material for training LLMs and generative AI models. Privacy will be protected. Interactions with Duet AI will remain confined to organisation.

AI pair programmers, what Google calls Duet, are being used to do what software engineers do not like to do. Software engineers are not fond of writing documentation. AI model can generate it. Engineers do not write code reviews. Some experts do it. It takes time. However, the AI model can do it. Some programme may be written in Cobol. An organisation may not have anybody who understands it. It is to be converted to a modern language like Java. This is legacy conversion. Duet can automatically convert it. All this does not mean that programmers are being replaced by the model. In fact, Google helps them go faster.

Milky way is home to 100 billion planets, attached to various suns or stars, and a couple of billion are similar to our own sun. In all probability, some of these planets may be home to civilizations more advanced than we are. As we know microbial life emerged on earth after it cooled down, and most star systems are billions of years older than our sun, and therefore life there could have emerged much before the emergence of single cell organisms formed here on earth.

There is research on extra-terrestrial life. It is possible that aliens or the hardware they have built might have visited the earth. There are astronomical oddities, reported sightings of UFOs or unidentified flying objects, now called UAP or unidentified anomalous phenomena. Even believers of God believe that there is extra-terrestrial life.

In October 2017, a telescope in Mauti detected something that has entered the solar system. Astronomers called it Oumuamua, meaning scout messenger. It was the first interstellar object that has been recorded. It seemed to cover a vast distance from another star system to enter our solar system. It was mysterious what its origin was. Its contours did not fit in other known astronomical categories. It could have been a rock, or an asteroid. As a rock, it was strange. Its length equalled the length of a football field. The shape was difficult to conceive — it looked like a long thin cigar. The more it was studied, the more its weird nature appeared and that struck the astronomers.

While travelling it approached the sun and then speeded up. Its acceleration could not be explained by the sun’s gravity alone. Maybe, it was a comet to have that extra acceleration. Comets being rocky snowballs, in proximity of the sun, the ice within them gets converted into water vapour. Gases are released to give them a speed boost, In addition, comets have a signature tail. In case of Oumuamua, no telescope observed its tail.

Avi Loeb, a theoretical astrophysicist at Harvard University, followed this news. He surmised that its acceleration could be attributed to sunlight. He compared its journey to a light sail, similar to wind sail of a boat. If Loeb is right, it means the object is not a natural phenonmeon, but an extra-terrestrial artifact.

Loeb is open to unconventional ideas. He published his Oumuamua hypothesis. He was saying as if an alien ship had arrived. Soon journalists made to his residence. Loeb is not alone in hypothesizing that universe could be filled with life. Astronomers too believe in this hypothesis. However, Loeb departs from them in thinking that aliens on other planets could have already made their way to us. He is outspoken about extra-terrestrial life.

In 2021, he wrote a book about Oumuamua called Extraterrestrial. It had become the best seller. His new book Intersteller will soon appear in the market.

Astronomers feel that Loeb should concentrate on new discoveries being made by James Web Space Telescope. Some feel he is not following scientific method. His disposition is speculative.

At the level of enterprise, AI makes sense when it is combined with the data of that company. AI drives value by working with the data.

Google has launched BigQuery Studio. It is a new way to deal with the data. Data provides new insights to the companies. Many organisations seek to derive value from investments made into big data and AI.

There are challenges in analyzing big data. Google, therefore, has launched BigQuery Studio. It is in fact a serverless data warehouse. It allows one to edit programing languages (SQL, Python and Spark) and ML at large scale (say petabyte scale). BigQuery Studio and AI people are the two sides to provide access to all services that the organisation needs. BigQuery Studio deals with the data — discovery, exploration, analysis and prediction of data. It could be started in a programming note book which later be opened in a programming note book which later can be opened in other services, say Vertex AI (ML platform). It allows teams to access data from where they are working. The controls are added.

Both generation and processing of data could be seen in BigQuery Studio.It then gets used in AI. The ML code can directly be infused in BigQuery as infrastructure.

BigQuery Studio facilitates the progression of an organisation adopting AI to cloud. It is a big market, and Google wants a sizeable market share.

Brad Smith, president of Microsoft, recently visited India. He expressed his views on India’s data protection law and AI. He appreciated that the data protection law has recognised data processor and data fiduciaries . Even data’s cross border movement has been allowed.

AI has the potential to do a lot of good. It is for us to do so responsibly and wisely. Machines may make decisions which so far were made by humans. We have to get it right.

AI may act on its own. That is an issue. AI should be a tool to serve people. AI products are coming too fast. There should be adequate usage of the product over a period of time.

AI laws should be built on existing laws. Non-IT persons should also develop AI skills. There should be collaboration with other countries. There should be focus on safety when a public agency uses AI. There has to be arrangement to slow down AI or turn it off when AI is used for infrastructure. It will ensure public safety. The control could be at the level of a model, or an application or at the level of a data centre, or at multiple layers.

Policy makers may be overwhelmingly optimistic or overwhelmingly concerned. Mostly they are balanced. They are testing the waters before acting.

2023 is the year of AI, and generative AI. It will stand out historically as an inflection point for the tech sector. Gen AI will be far more transformative for people’s lives than perhaps any other invention in people’s lifetime.

Brad Smith compares ChatGPT to printing press. Gutenberg perfected the printing press in 1462 using movable-type. Printing press made knowledge accessible to more people. Generative AI is similar. It is a tool to learn and to do research.

AI’s costing causes some concern. However, these are early days. These days everybody deals with LLMs or large language models, say GPT-4. In future, we may shift to narrower models, say some open source models which are smaller. These may not be as efficient as the larger models in one or two tasks. However, at the same time, they may do as well as the large model in one or two tasks. There will be constant innovation. The large models consume more energy. Companies should optimise the GPUs so as to make models more efficient computationally. There should be more investment in green energy.

IBM CEO and MD suggested to the government to develop sovereign capability in AI and to set up a national AI computing centre. There is constant talk about the capability of AI and the need to regulate it. However, for the first time, we are talking about the government developing AI capabilities. The government can make use of datasets and LLMs for macro-economic purposes.

As we know, vast datasets are required to train LLMs. These models are basically predictive. If the government uses appropriate language models, the government can use these to predict monsoon, the health requirements of the citizens over the years, the way the economy and GDP would grow. These areas are of no interest to private organisations. If the government creates AI capability centre, it could have the right kind of AI tools for policy making.

The government will have server farms and would do appropriate investment in them. These server farms will be housed in huge real estates. To develop such infrastructure, the government can allow private participation.

Advertising Standards Council of India (ASCI) has issued guidelines for financial and health influencers.

According to these new guidelines, influencers providing financial advice must prominently display their SEBI registration number as well as qualifications in adverts and on social media platforms. Appropriate credentials from SEBI, IRDAI, ICAI or ICSI have to be provided based on the nature of advice.

Health or nutrition influencers must disclose relevant qualifications such as medical degree or certification in nursing, nutrition, dietetics, physiotherapy or psychology etc.

In case of videos, the credentials must be superimposed on visuals prominently or upfront or mentioned in the opening remark. For blogs and text-based posts, they should be stated at the beginning before the consumer begins reading.

In case of audio-only media (including podcasts), they should be announced before the advertised content goes on air.

AI revolution is coming but not soon enough. It is seen as an exciting new wave of technology. There are issues of the use of data, the accuracy of AI-generated answers and the leakage of confidential data. AI, no doubt, will boost productivity. Yet, as we know, historically all inventions take time to have full adoption. Steam power and internet illustrate this. In 1990s, internet took the world by storm. It was predicted it would disrupt advertising, retailing and media. Though these predictions proved true, it almost took a decade. Over a period of time, the tech was refined and costs dropped. Broadband internet connections became commonplace. The payment gateways developed. Audio and video streaming developed. The internet adoption was funded by many investors who believed the startups, and things moved on.

AI could lead to a similar gold rush. There is an investment frenzy. Billions of dollars are pouring into generative AI. Corporates are sampling the technologies. It is to be seen how the industry gets affected. The advent of ChatGPT in Nov.2022 was a ‘netscape moment’ ( reference to the introduction of browser in 1994). ChatGPT has poured soul into internet. However, it is just a start. It has opened door for new opportunities. The widespread use of generative AI apps will take time. There is already improvement in currently known technology. However, future breakthroughs are yet awaited. Mainstream adoption could take 8 to 27 years. It is a broad range. It is going to be governed by economic cycles, government regulation, organisation culture (OC), and management decisions.

Here people are involved. We are not dealing with laws of natural sciences. There are human choices involved. Tech diffuses through people AI forecasts see a productivity surge and addition of trillions of dollars to the global economy.

Specialised AI refers to AI systems that are designed and trained for a specific task or a set of tasks. These AI systems are tailored to excel in domain and have limited capabilities beyond that domain. To illustrate, language translation systems, image recognition algorithms and recommendation engines.

Though generative AI is booming, it can be very costly. There is significant expense in training and executing LLMs. To begin with, early adopters of specialised AI solutions are companies in retail, healthcare, BFSI, and manufacturing. As the potential of specialised AI continues to unfold, its applicability holds promise for other industries.

Nations have to harness the power of AI to raise the global and their own GDP. India has been blessed with skilled manpower, expanding tech industry and a pool of data. The whole environment is conductive for economic transformation. Policy makers must recognise this opportunity.

India has demographic dividend advantage along with technological advantage. This creates multiplier effect. Global economics is moving towards AI-driven industries. AI must be adopted by India not merely as a tool but as the foundation of its economic strategy. India should not confine itself to being a mere consumer of AI but should be a creator of AI. India should seek to become an AI superpower.

Some countries rely on anonymised datasets and are involved in litigation. Instead, there should be dependence on computational privacy guarantees in the use of aggregate data sets. The AI systems should be safe. They must be based on reliable and traceable datasets. India should create an ecosystem of AI modellers.

AI can help in addressing the social and economic disparities. It can help in optimising resource allocation. It can create job opportunities.

India should create the hardware infrastructure of computational tensor units. There should be enough GPU units. The government’s role here is very vital.

If India neglects AI as a strategic tool, and remains just a consumer of AI technology, there will be missed opportunities, economic stagnation and loss of global influence.