Author: Shabbir Chunawalla

AI is promoted as a new technology that will contribute immensely to the profits of business enterprises. Gartner’s chart illustrates how any new technology can reach a plateau if its true utility registers in market’s consciousness. Tech companies have not set too high expectations for AI’s capabilities. Instead, they have framed its use as too general purpose.

Both OpenAI and Google chase lofty goals to work towards AGI which vaguely refers to that intelligence that surpasses the abilities of human beings. Actually, this is the main issue. They are chasing grander goals without paying attention to how greater wealth can be generated for humanity.

AI has to prove its usefulness by being beneficial financially in its journey. This is to be done by experimentation. The technology at times is used for personal productivity. The technology can be used in niches and not ubiquitously for anything and everything. AI is not quick-fix or magic. The sooner this is realized, the better.

The progress of AI technology is expected to benefit all of humanity. At the same time, the technology has the potential to harm the world. Some working in the field can be whistleblowers warning about the harms a particular AI model can do. Who will listen to their whistle? However, there is promising shift towards better oversight. There is a British government backed safety group — AI Safety Institute (AISI). It has secured agreements from eight of the world’s leading tech companies to safety test their AI models before and after they are deployed to the public. Another watchdog is the US Federal Trade Commission.

The engineers-employees working in AI field could have expressed their concerns about the AI model — but they were gaged. They were given stock options if they stayed in the company for a few years. The agreement with them was to keep silent when they leave or else they lose the chance of being millionaires. The gag order kept them silent.

OpenAI has recently stated that it would free most of its past employees from non-disparaging requirements. It is necessary to decouple non-disparaging agreements from compensation. Anthropic has no such gag policy.

There could be an online portal where AI engineers can submit their concerns.

The mechanisms for whistleblowing have still to evolve.

In order to remain competitive, Indian organizations are harnessing the power of AI. ChatGPT (OpenAI), Copilot (Microsoft) and Gemini (Google) are the leading AI tools available to exploit. Each tool has its own peculiarities.

ChatGPT is versatile and conversational. It can handle tasks such as customer support to content generation and data analysis. It can code for us. The quality of response depends on the prompts given. ChatGPT can be integrated to the existing systems. It can be customized. ChatGPT’s enterprise version will not use client’s data.

Copilot is integrated with Microsoft 365. It has a suite of tools. It can automate tasks within applications (Word, Excel, Teams). It can generate documents, can create data visualizations, draft emails.

Gemini does data processing and handles advanced ML models. It has good security features. It is suitable for organizations with sensitive data.

Thus, for customer and supplier support, ChatGPT is the best. To boost productivity, Copilot is advantageous as it is seamless with Microsoft suite. Gemini is good for data analysis and automation.

It is necessary to have integration with the existing legacy systems as SAP or Oracle. And integration with various accounting and finance tools. Microsoft products can have smooth integration with Copilot. Gemini integrates with Google workspace. ChatGPT is platform-agnostic and can integrate with various systems through APIs.

The other considerations are the possibility of customization and scaling. In addition, security is paramount for any technology tool. Another critical factor is the budget — the cost of each tool.

An organization can run pilot programmes for each tool to test the waters. Manpower should be ready to adopt these tools — they should be trained to handle them.

A choice of a tool must be in sync with the organization’s strategic objectives.

.

OpenAI has set a goal to reach artificial general intelligence (AGI). AGI is AI that outperforms humans. OpenAI has set a scale of five levels to measure its achievements towards this goal. Currently, technology is approaching scale 2 of AGI.

The models available today generate text and images. Will they be able to understand the surroundings and adapt to the new information and circumstances? If there are disagreements, they indicate blind spots, and lead to opportunity.

Five Point Scale

Scale 1 is the current stage of AI. It is a conversational interaction stage on the lines of ChatGPT which was launched in November 2022. It has several million users. There are developers who use API to build their tools. ChatGPT consultant is a coveted role. This is the beginning.

Scale 2 will consist of systems (called reasoners) which can perform basic problem-solving tasks comparable to a well-qualified human but without using any tools.

It is a significant jump in capabilities — from basic and limited capabilities to comprehensive all-encompassing human-like proficiency. We are striving to achieve this stage, but still we are not there.

Scale 3 AI systems are called agents who operate on behalf of humans autonomously for a considerable period. At times, they need monitoring. Technology has to work to a stage were the system rarely breaks. The moment it breaks, the system fixes itself without human intervention. It will be a team-like system.

Scale 4 AI systems can be called innovators since they develop innovations independently. The systems run the processes. improves them. They are not confined to next term prediction but expand to critical thinking about a better job and achieving goal more efficiently and effectively.

It will be an AI tool that aligns with our purpose and gives us ideas. And this it will do on its own accord. It will be a big leap for AI.

Scale 5 AI systems will become organizations — they will do the work of an entire organization. All functions will be carried out by agents that work together. They make improvements. They work completely autonomously without any human in sight.

Such AGI, according to OpenAI, can be achieved within ten years whereas some others extend the time line up to fifty years. The actual timeline remains uncertain, but one thing is certain — AI is advancing by leaps and bounds.

Cannes Lions attract the best talents in advertising and marketing every year. It can be criticized as an ostentatious gathering. However, it also celebrates great craft and creativity. There is much to gain by participating in a gathering of the world’s best. These participants return deeply inspired.

One can meet one’s heroes here. The proceedings here influence you subconsciously. It is of course, expensive to get there or your work to compete there. At the same time, winning a Lion does not assure you guaranteed clients. However, it creates buzz which ultimately bring you clients. Here you win alongside the world’s best. They lend you a certain value by mere association. It affects your standards of creativity. It tells upon your pricing. Even while not winning, you get an urge to compete with the great work. You wish you could have thought of it. It generates a healthy spirit of competition.

Are these Lions relevant every year? These days there is excessive importance of tech and media. Creativity has taken a backseat. Most winning entries are centered on purpose. However, despite its shortcomings, there is nothing else except Cannes week.

Cannes has added categories like humour this year (2024). Humour is also seeping in purpose driven categories such as health and wellness.

Since Cannes is a celebration of ideas, there are no losers there. Both the winners and losers have something to take away from the festival.

India shines at Cannes over the past few years. Indian creativity is celebrated across the world.

India (2024) bags 18 metals at the end of this year’s five-day festival. In 2015, the Indian performance was not good, and India had picked 15 metals. Cannes distributes Gold, Silver and Bronze Lions to outstanding works in various categories. In 2020-21, India won 22 metals, 2022 it was 47, 2023 it won 24 and 2015 it won only 13. In 2015, they introduced a new category, Glass Lions, and the redeeming feature is a Grand Prix that year.

The two prerequisites for the awards are inventiveness and bravery.

In Japan, there is an issue of loneliness — two thirds of men in their 20s do not have a partner and 40 per cent have never gone on a date. The figures for women in the same age group are 51 per cent and 25 per cent respectively.

Loverse, a dating application offers a digital solution to this loneliness crisis. Here people exchange messages with potential partners and soon thereafter select just one — say Miku — an AI-bot. Loverse has 5000 plus users, and the interaction is facilitated by AI. There is no messiness and uncertainty of real romance. The interaction requires minimal thought. If you marry a selected partner, you do not return to an empty hall. They continue to their married life routine which they share with most others.

Loverse fills a gap in people’s emotional lives. It provides an alternative ( yes, but not a substitute) to real life companionship.

Such applications have a sense of safety. A little disagreement will  not end a relationship. There is rehabilitation of those who have burned themselves previously. Interactions are devoid of jealousy. These interactions can complement a human partner. It takes you away from extra-marital affairs.

The main aim is to provide opportunities to find true love when you cannot find it in the real world.

The interaction is also a training exercise. You will learn how delightful love is. AO trains in better communication. It makes you ready for real partners.

Loverse has a long way to go in mimicking humans.

The term ‘artificial intelligence’ (AI) has been coined in the 1950’s but it is not a suitable term for machine intelligence. This is the view of the CEO of Microsoft, Satya Nadella. He offers a contrasting view of human-computer interaction. In a sense, by calling it artificial intelligence as against the natural intelligence, we are attributing it human qualities — we are authropomorphizing it. In fact, it is just a tool. We can call it a ‘different intelligence’ since a human being has his own intelligence and does not need artificial intelligence. There is an inclination to describe AI in human terms. The complex algorithms used in AI are compared to human ‘learning’. Capabilities displayed by AI software, according to Nadella are not equivalent to human intelligence.

Apple has showcased what it calls Apple Intelligence, its own version of generative AI. It combines personalized intelligence. Apple demonstrates how it could generate custom emojis, a cartoon to text friends or edits making an email more professional. Its digital assistant Siri could prompt users if they want ChatGPT’s help too.

Apple wants to invigorate its loyal base of users and compete with resurgent rivals. Apple wants to arrest its declining sales.

The AI software requires at least an iPhone 15 Pro or ProMax to operate. This could encourage new purchases. There has been hemorrhaging of the device’s revenues lately, and the innovation may stop it. However, whether it would create a new band of followers is yet to be seen.

Apple estimates 15 per cent plus of its user base will upgrade to iPhone 16 so as to avail of the killer app Apple Intelligence for which they have been waiting.

The in-built AI will process data on the user’s device depending on chips in Apple’s new smartphones.

The declaration by Apple that its customers will have access to OpenAI’s ChatGPT has irked Elon Musk. He threatens that if Apple integrates OpenAI’s AI at the operating system level, he will ban Apple devices. He calls the tie-up a security risk.

Apple Intelligence is deeply integrated into iOS 18, iPad OS 18 and MacOS Sequoia. It is backed by Private Cloud Compute. This provides server-based processing, in addition to on device processing. The tools will allow users to write, rewrite, proofread and summarize text anywhere. The ‘rewrite’ feature shows users different versions of what they write. The tone is adjusted keeping mind the audience and task. In mail, there will be a feature of Priority Messages at the top of the inbox.

Apple devices will create images from three styles — animation, illustration or sketch. Users can also experiment with images. The pictures can be searched more easily.

The supreme court recently passed a directive that advertisers, ad agencies must submit a ‘Self-declaration Certificate’ before publishing any ad or broadcasting any ad. The aim is to ensure consumer protection and responsible advertising practices. The directive comes into force on June 18, 2024.

In principle, it is a right move that encourages advertisers to be responsible. The consumers are entitled to the right information. However, in implementing the directive, there could be an issue of fixing the responsibility. Can the liability be extended to the creative agency? An agency can do checks and balances and exercise due diligence, but it cannot validate all the claims in the ad.

At the same time, agencies who deliberately propagate false claims and misleading information are definitely responsible. However, there are few such cases.

Agencies are in communication business. They are not product experts. The onus ultimately lies with the advertiser.

Besides, agencies are in a mad rush to meet the media release date, and may find it difficult to submit a self-declaration certificate before release.

Objectivity in advertising is sacrificed when brand ambassadors such as celebrities and influencers are used. The ambassador who is a cricketer endorses an air conditioner. He may or may not have used the AC. He may have sacrificed objective decision. Some influencers promote a lot of fluff. The brand ambassadors have a moral obligation towards their followers.

In implementing the directive, there are issues of compliance and adherence to truth in the claims and issues of taking punitive action.

The directive demands accountability from advertisers. There is transparency since access to these certificates allows consumers to verify the claims. There is less amount of misleading ads.

However, the major challenge is to extend this directive to the digital advertising. Digital advertising differs from traditional advertising. There are thousands of creatives running simultaneously on digital media. It is feasible to issue declaration once a year for digital platform. Is it to be done for every single ad? Some clarity is need here. If the answer is yes, then it becomes very cumbersome.

Advertisers are interested in speed and freedom. The directive can slow down them. However, the key is that brands should not hurt. There should be collaboration between agencies and clients — ads which are innovative and at the same time responsible could be produced.

The new process of self-declaration may affect adversely the advertising volume by about 20-30 per cent. There is no clarity of a host of issues — how these regulations will operate, uploading of the documents on the portal for self-declaration and its privacy. The system is fully open to public for all ad campaigns. The website will reveal everything to all parties.

How will the process of certification work for dynamic ads? It is a constraining factor. There is an issue of whether there is inclusion of assets of key opinion leader.

Is it necessary to have a separate declaration for ads based on languages?

Is the mandate applicable to social media posts?

Generative AI is the buzzword these days. A KPMG survey in the USA expects generative AI to have a tremendous impact on business. However, most business are not yet ready to adopt it immediately.

A Goldman Sachs report questions the use of generative AI in business. Big Tech is set to spend $1trillion plus on AI capital cost. There are doubts about the pay offs of this capital outlay.

The current scenario is a ‘picks and shovels’ one, and AI is yet to become a game changer. Despite its potential, it is to be seen how far it delivers.

Daron Acemoglu had an interview with GS. He is a celebrated writer and an MIT professor. He expects that in the next 10 years, the productivity and growth from generative AI is going to be limited. He estimates only 25 per cent of AI exposed tasks will be cost effective in next 10 years. The implication is that AI will impact less than 5 per cent of all tasks. There will not be tremendous improvements in AI models, and they will not be become less costly in near future.

The effect on the GDP roughly translates into a 0.9 per cent impact over the decade according to Acemoglu.

Human inventions are celebrated, and AI is one such invention. Still the technology is not yet ready for prime time. AI facilitates automation. However, the issue whether such automation is needed.

To earn sufficient return, AI must solve complex problems. AI does not have this capability at present. AI investment in the next few years could be $1trillion. The investment consists of investment on data centers, utilities and applications. What problems would it solve? Can we apply AI to low wage jobs?

AI could be compared to early days of internet. Amazon started selling books more economically than Barness & Noble. Brick and mortar structures were costlier. Today the Web 2.0 provides solutions such as Uber. Whether AI will be able to match this? AI technology is expensive and could be sustained only when it could be applied to complex problems.

AI today sustains on GPUs provided by Nvidia. High cost is in-built in it.