As we live through the AI tidal revolution, there is both excitement and anxiety about what it can mean to livelihoods and jobs as we currently know them. Many of the world’s premium firms now use AI to screen applicants for customer engagement, and more. If you are a young professional or entrepreneur, you likely are or your organisation is already using AI in some manner. We at TAQ, sought to understand how you can best leverage AI in workplaces, decision-making, workflows, etc. In this article, we look at the 5 bold predictions AI scientist Dr Rana el Kaliouby made in a recent podcast episode of Rapid Response for where AI is going. And propose some applications and tools to prepare for that future both to enhance your work and build productivity and wellness into your day. We aim to be tech-optimists, and the numbers reflecting how AI is benefitting businesses in India help reinforce our belief. 78 per cent of small and medium-sized businesses have reported revenue growth because of AI adoption; another report suggests India is leading in AI adoption at 30 per cent, surpassing the global average of 26 per cent.
Dr Rana makes 5 predictions for where AI will go next:
1. Emergence of Agentic AI
Agentic AI is an independently acting AI that, unlike traditional AI systems, does not need human intervention to make decisions. Based on its objectives, it learns from experiences, gathers data from its environment, and makes decisions that help the goals it is programmed to achieve. For instance, an e-commerce platform could deploy an AI agent to learn about purchasing behaviour and the usual issues its customers face with the products. The agent will accordingly tailor recommendations, predict the issues, and pre-emptively solve them.
AI such as UiPath—which executes repetitive tasks, and IBM Watson which provides decision-support in multiple sectors through large datasets, are great examples of agentic AI that can be adopted at an organisational scale. To manage work at a personal level, you can utilise tools like Todoist (to prioritise tasks), and Focus@Will (to avoid distractions).
2. Prevalence of Embodied AI in our workplaces and homes
Embodied AI systems marry features of generative AI with physical entities such as robots and drones. They can have sensors like optical vision and cameras to familiarise them with the environment and function autonomously. Imagine a robot with heavy-lifting equipment on a manufacturing floor or a construction site, or a robotic arm that does intrinsic surgery—this is embodied AI. How easy it becomes for farmers—performing one of the most labour-intensive tasks–to use drones to analyse their crops’ health, production, and yield? In fact, embodied AI is being adopted not just for work, but for leisure and self-care at the workplace as well. Clockwork created the world’s first AI-powered robot for manicures that can be stationed at workplaces—with the idea that you’ve 10 minutes at work—you can get your nails painted in that time.
3. Companies that are one-person unicorn
Sam Altman, the CEO of OpenAI believes that AI will allow an organisation to surpass a billion-dollar valuation without hiring anyone. These ‘one-person unicorns’ will be possible when AI modules perform all the admin tasks for an organisation.
But until that time comes, companies can currently use agentic AIs that go several steps further than large language models and are able to complete complex tasks. Relevance AI is one such example that can assist with sales, customer support, research, marketing, and operations.
As an early entrepreneur, such tools can be immensely beneficial for scaling with limited resources. You may have a strong sales team, but your marketing leaves something to be desired—explore how that task can be taken up by AI. That is the benefit of the dynamic working modules of AI—the foundation is already laid; all it needs is to tweak it as per your requirements.
4. AI health co-pilots
India has a serious case of worker burnout. Mental health chatbots that provide personalised responses can be beneficial in keeping records of emotional well-being, provide immediate support, and even act as assistants to mental health professionals. You can use these to detect early signs of stress and burnout and provide timely interventions.
Access to a personal mini-therapist on your phone, or even a doctor for that matter that can analyse your symptoms for free is becoming a reality already. But we do know what putting our symptoms in Google does—to avoid the spread of misinformation and panic, it is important that AI provides accuracy and uses the data responsibly. Wysa and Woebot are some mental health chatbots that have proven to be effective in the workplace. Noom, Ultrahuman, MyPlate and Activ can track your biomarkers continuously and help you make good activity and food choices.
5. Emotion AI
Emotion AI can detect, interpret, and respond to human emotions in real-time. It gathers and reacts according to voice tone, body language, heart rate, and more. In a workplace, such AI can be leveraged to effectively respond to customers, and employee well-being, and manage team morale in an empathetic manner, instead of providing standard responses. Affectiva is a great example of emotion AI. The one drawback is of how to protect the misuse of people’s information from Emotion AI. But familiarity with where this space is going is useful. In recruiting, bots are already being leveraged for early-stage job interviews. Knowing how emotions are detected and read must be part of a young professional’s interview prep.
Ethics and sustainability cannot be ignored
What undercuts all the AI use cases, are the ethical and environmental implications. AI is trained on preexisting data, so it can replicate biases and blind spots present in the original dataset. Even if organisations provide their own datasets to the AI modules to tailor the technology as per their own need, it still raises massive privacy and misuse concerns. Over time, companies must develop internal governance mechanisms to ensure information isn’t misused, consent is obtained, and keep regular checks on how the AI is functioning.
Similarly, training complex AI models lead to a staggering carbon footprint— According to researchers like Shaolei Ren at the University of California, Riverside, the water consumption required to handle just 5-50 prompts on Chat GPT is around 500 millilitres per interaction. So, even as companies develop more energy-efficient algorithms that require less computational power, developing personal discipline about the use of Chat GPT is also important to make the application of AI sustainable.