Author: Shabbir Chunawalla

OpenAI’s intends to realize the goal of AGI — the point at which AI capabilities supercede those of a human being. This is a point of contention — the timeline prediction ranges from a decade to eternity — not foreseeable in near future. Even AGI has not been defined precisely.

Investments of billions of dollars is pouring into the infrastructure for AI, and the stakes are high as we may be many years away from pay-back.

Sam Altman is optimistic. In this week (November 2024), at Reddit AMA, he claimed that AGI is achievable with current hardware. It is expected of a CEO of a major AI company who is committed to achieve AGI. Still, even by Altman’s standards, it is an overambitious statement. It simply means that the current capital outlay of OpenAI will soon pay back. Still, it is not clear what he means by the current hardware.

It is tantamount to saying by Tesla a decade bulk that the company would place an autonomous car, and still the best of Tesla’s self-driving cars needs to be monitored by a human driver.

It is in the interest of OpenAI to claim that the present capital outlay ensures a bright future, and builds up the valuation of the company, and elicits investor interest.

It is audacious for Altman to claim that superintelligence is a mere ‘ a few thousand days’ away. It is so frivolous. Of course, it will take longer, but he is confident he will reach there.

It is not difficult to visualize a future where machines defeat us at video games or script a movie. However, machines may go much farther and can surpass us the ways we have not imagined. It is not a utopian idea but is going to be a reality in our lifetime.

Of course, we have AI systems today. Still, they are not as advanced as the human brain. Though they can perform specific tasks, they lack the broad understanding and adaptability which characterizes human intelligence. We are advancing towards artificial general intelligence (AGI) which may remove these deficiencies and may match human-level capability of cognition. Going beyond AGI, the next stage is going to artificial superintelligence (ASI). In fact, ASI would redefine human existence.

Human intelligence is constrained by biology, whereas ASI will operate at digital speeds. It will solve huge problems millions of times faster. Can you imagine something that glances over all the research papers on an issue and bring forward a solution while we are busy doing routine things? It could be called intelligence explosion — AI system faster than humans and cannot matched or controlled.

Its impact is breathtaking and profound. It will revolutionize healthcare, quantum physics and climate change. However, the same plus points could threaten our existence if it is not aligned properly with values and interests. An ASI system dealing with cancer can decide that making living things extinct is the solution to get rid of the cancer. Such logic is unpredictable and esoteric.

ASI challenge is not confined to technical aspects only. It is a race to control what we are creating. There are issues of ethics, governance and human intervention.

We have to align the system that is capable of rewriting its own code.

Superintelligence journey is not pre-decided but is not avoidable. The question now is not whether we should develop ASI, but how to prepare once it arrives. There should be investment in AI security research and international discussion.

It is the most significant technological leap in the history. It has the potential to be the biggest achievement or humanity’s last invention. The challenge is not merely technical. It is philosophical, human and ethical. ASI should benefit us, rather than replace us.

.

Once upon a time, Nokia was the market leader in cell phones. As soon as double SIM phones and smart phones stormed the market, Nokia was displaced from the market. It was unprepared for the changes in the market.

Singularity is a period of rapid technological change with a deep impact, transforming human life and business irreversibly — (Kurzweil). There are mind-blowing changes all around — business, society, climate and geopolitics.

One by one, industry will experience exponential change. There would be a sudden change in customer preferences. Technological advance will be so rapid. The manpower required and the skill required will be different. Many organizations will be caught napping. It is the Singularity future. The most crucial element will be the rapid change. The current 500 large corporations will no longer remain large, some 40 per cent new companies will replace them in near future.

Intel has lagged behind in the race of AI. Nvidia who used to make gaming chips diversified into AI chips and have become the market leader.

Boeing was a coveted corporation in aerospace area. Boeing Max faced crashes, as pilots were not trained on new software. Its Starliner faced hitches in space and was brought back to the earth without its crew of astronauts.

In the next decade, only those organizations will prosper whose leadership is able to deal with the pace of change. Either you disrupt or be disrupted.

Masayoshi Son, CEO, Softbank Group, expressed his confidence in the coming artificial super intelligence (ASI), and said that this technology will require hundreds of billions of dollars investment to realize.

According to his prediction, generative AI will require $900 trillion in cumulative capital expenditure in data centers and chips. He feels that Nvidia, the chip maker has been undervalued on this basis.

ASI will be 10,000 times smarter than a human brain and will exist by 2035. He made this prediction at a conference in Riyadh, Saudi Arabia.

Son is saving up funds so that he can make the next big move. His record as an investor is patchy, e.g. Vision Fund, 2017 startups have crashed in value. Saudi Arabia’s sovereign wealth fund has invested $45 billion into the first Vision Fund.

Organizations expect a productivity boost ranging from 10 to 40 per cent by introducing AI, including generative AI. It is imperative for the companies to become AI ready for this. Very few organisations are ready for AI adoption.

The key requirements for AI adoption are — talent, strategy, governance, data and technology.

The pre-requisites for AI implementation are ML capabilities, pre-built algorithms and dynamic compute.

The data accuracy and accessibity are significant challenges.

There should be AI strategy that aligns business objectives and strategy.

The governance part consists of management of risks such as bias, misuse and security threats.

There is a need to upskill the workforce.

It is necessary to have a scalable and robust foundation. This must be supplemented by an AI foundry and factory model to scale up AI initiatives across the enterprise.

Generative AI has affected visual representation — it has become dynamic and tailored. DALL-E and Midjourney create unique graphics based on learned patterns. It quickens production and makes visuals personalised.

With AI capabilities, fashion firms create virtual models. The commercials shown are tailored to target audiences. That enhances the conversion rates and loyalty.

Owing to generative AI, creatives are made faster. These make campaigns adaptable. The visuals can be modified in real time in response to consumer feedback and trends prevailing.

Fashion industry has been affected — there are provisions of AR features and virtual try-ons. Interactivity affects buying experience. Customers can try on clothing digitally in virtual fitting rooms.

Generative AI also affects visual storytelling. Marketers can create immersive stories.

As AI develops, we can have virtual and real-world graphics. Interactive Flat Panels (IFP) and generative AI are synergistic. There would be tailored presentations. AI can be integrated with technologies like AR and VR to give immersive experiences.

In India, there are 13 million small grocery shops, and over 10 million of them are in tier-2 and smaller cities. Since the rise of quick commerce organizations such as Zepto, Blinkit and Swiggy, nearly 2 lacs grocery shops have closed. As many as 60,000 stores have closed in the tier-1 cities, and an additional 50,000 have shut down in tier-2 and tier-3 cities.

This festive season the sales of grocery shops have remained stagnant. Even customer visits to grocery shops are down by 50 per cent.

Consumer goods companies have said that the demand for their products have increased on quick commerce platforms. The consumer preferences have shifted .

The smaller shops struggle to match discounts offered by online and quick commerce platforms. It puts pressure on their margins.

Quick delivery, suitably wrapped and packed commodities, deep discounts have created an unfair playing field which erodes the customer base and profitability of smaller grocery shops. Along with these aggressive practices, there is overall economic slowdown. It has led to the closure of many small retail shops.

MBA has a history of more than 100 years, since the start of the first MBA in 1908 by University Graduate School of Administration (now Harvard Business School). MBA was created in response to the growing demand from business for a formal course for business managers required to man the growing industrialisation in the US.

The pioneering MBA course was of two years and attracted 80 students who were taught by 15 faculty members.

MBA has been evolving since then. It has a different format in different countries. Today, it again calls for a big change. Its core has remained the same for the last 50 or 75 years.

We spell out the challenges it faces, so as to know how to revamp it.

1. To begin with, the course emphasized a combination of soft skills and hard skills. These days there is a need for more hard skills.

This is seen in the emergence of degrees such as master’s in finance, master’s in data science and so on.

2. It is believed that the added value from MBA comes from critical thinking and analysis. There is also the ability to synthesize and connect the dots. The inputs of basic accounts, 4Ps of marketing, MBO etc. are alright and can be mastered.

3. The entry level MBAs do not match the industry requirements. The MBAs from elite institutions have an advantage as they undergo a rigorous selection process and have a higher perceived caliber.

4. Apart from analytical and synthetic skills, industry also require skills such as data visualization, programming in R and Python, Excel, PowerPoint and so on. It is difficult to provide all these skills in a single programme.

5. Each business graduate must be digitally and technologically savvy. There should be courses on AI, cybersecurity, geopolitical environment, government and business and so on. The distance between business school and public administration schools is reducing.

6. Residential business schools should allow access to recruiters and social network.

The business schools should adopt case-driven method where students are encouraged to be inside in business scenario to find a solution to the problem. The students are in driver’s seats as they tackle the problem. The faculty members simply ask the questions and facilitate the discussion.

The revenues of film studios keep fluctuating demonstrating the fickleness of this business. In 2023, there were four films with earnings over Rs.500 crore at the box office. It lifted the spirits of the industry struggling for footfalls into the theatres post pandemic lull and a shift towards digital.

However, in 2024, there is a tepid response with just one movie going past Rs.500 crore club so far. Even in 2023, it was the higher prices of tickets that boosted revenues rather than the increased footfalls.

The tried and tested formulae do not seem to be working. Movies do require better quality of writing and directing. A diverse bouquet of content would satisfy the audience. While studios corporatize and bring in corporate capital, they should also improve the quality of movies being produced.

Karan Johar’s Dharma Productions is a family-founded studio. A vaccine maker Adar Poonawalla made a strategic investment in this firm of Rs. 1000 crore and got 50% stake in the firm. Studios are a capital intensive businesses which require capital to make films. They also hedge their bels on streaming platforms and TV. Dharma Productions will get more autonomy and creative control. There is consolidation at the global level too. In India too, there is Reliance-Disney and PVR-Inox consolidation.

Streaming giants (Netflix and Amazon Prime) are the biggest buyers of studio-driven contest.

Hindi film industry is experiencing stress. There are no regular releases and there are declining revenues. Dharma Productions revenue halved from Rs.1000 crore plus in 2023 to Rs.500 crore plus in 2024. YRF or Yash Raj Films too had revenues Rs.1500 crore in 2023, boosted by Pathan’s release. It was a huge surge. In 2024, it may close at Rs.700-750 crores.

AI plays a vital role in content creation. A TV serial’s full episode can be created using AI, and it can provide immersive interactive experience. AI can generate dialogues. It can help scriptwriters to perfect their scripts. Scriptwriters thus get more time to think things through. Thus, generative AI is a great enabler.

AI can reduce cost of production. It can diversify content types and devise now forms of content from scratch. Still storytelling is the core of scripting and requires the human touch.

AI can be integrated to content production. It is both time saving and cost saving. AI has moved from generating scripts to characters alones. Google’s NotebookM converts text into podcast-style audio formats. Character personas are tailored to specific audience preferences. Most of the use of AI is confined to pre-production stage. It is used for ideation and pre-visualisation.

AI in content production brings its own challenges — there are ethical and legal issues. There should be checks and balances in place. AI generated content could be distinguished by a tag. There should be quality control. To achieve QC, human intervention is necessary.

Nandan Nilkeni’s Fundamentum Partnership backed KuKu FM is audio content platform. It wants generative AI to become part of content creation, production and dissemination.

Generative AI models in scripting perform diverse functions. Some models handle science fiction well while others do not handle true crime well. The capability of models varies from genre to genre.