Here’s To An AI-Filled New Year

Welcome to Beyond Tomorrow!

We are in week 5 of our new weekly newsletter, crafted to keep you at the forefront of AI advancement. Every week, we’ll bring you a round-up of the most impactful AI stories, highlight how they shape our world, and provide insights into what’s next in the ever-evolving AI landscape.

If you’d like to receive Beyond Tomorrow in your inbox every week, click here to subscribe.

Here’s to an AI-filled New Year

As 2024 draws to a close and we look forward to 2025 by making New Year resolutions—exercising more, eating fewer sweets, spending more time with family, or whatever it is you wish to achieve in 2025, I wish you much success. Just one suggestion: Getting upskilled for artificial intelligence would be an apt New Year resolution for our times, because we all need to be in step with AI.

But whether you keep your resolutions or not, one thing is certain and there won't be any failures on this score: AI will continue to become even more important and integral to our lives and work in 2025. The AI juggernaut will not stop and as businesses and smart folks across the world realised a while ago, you either ride it or risk getting crushed under the changes AI will bring.   
  
I have my own list of what I'm looking forward to in 2025 and I have written on it in this newsletter earlier, so I won't go down that path. Rather, I decided to quiz Google's Gemini 2.0 about the major developments in AI expected in 2025. Here's what Gemini 2.0 had to say (edited for brevity):

• Continued Advance of Multimodal Capabilities: Expect models that handle complex multimedia interactions even more seamlessly.   

• We might see more sophisticated video understanding and generation, allowing AI to "watch" and analyse videos in real-time.

• More advanced integration of different modalities, leading to AI systems that can reason and create across text, images, audio, and video in a more unified way.   

• More Powerful and Autonomous AI Agents: The trend of AI agents will likely accelerate.

• Agents might become capable of handling more complex, multi-step tasks with increased autonomy, requiring less human intervention.   

• We might see greater collaboration between multiple AI agents, working together to solve complex problems.

• Enterprise adoption of agents will likely increase, impacting various industries.

• Emphasis on Explainability and Interpretability: As AI systems become more complex, the need to understand their decision-making processes will grow.

• More research on making AI "explainable" (XAI) to increase trust and transparency.   

• Development of tools and techniques to help humans interpret how AI models arrive at their conclusions.

• Edge AI and TinyML Expansion: On-device AI and smaller, more efficient models will continue to gain traction.

• More powerful AI capabilities on smartphones, wearables, and IoT devices.   

• TinyML (machine learning for microcontrollers) will likely expand into new applications, like industrial automation and environmental monitoring.   

• AI for Scientific Discovery: AI's role in scientific research will likely expand.

• AI could be used to accelerate drug discovery, materials science, climate modelling, and other fields.   

• We might see AI systems that can formulate hypotheses, design experiments, and analyse data with greater autonomy.

• Deepfakes and Synthetic Media Pose Challenges: While AI can be used for good, the ability to create realistic synthetic media will pose increasing challenges in areas like disinformation and fraud.   

• More advanced deepfake detection methods will be needed.

• Ethical and legal frameworks will need to evolve to address the implications of synthetic media.

• Continued Debate on AI Safety and Regulation: The conversation around AI ethics, safety, and responsible development will continue to intensify.

• More detailed regulations may be implemented, impacting development and deployment of AI systems.

• Greater emphasis on aligning AI with human values and ensuring it benefits society as a whole.

We'll circle back sometime next year to see if Gemini 2.0 got predictions right. Meanwhile here are some interesting stories on AI from NDTV Profit:

The Reserve Bank of India has set up an eight-member committee to draft an AI framework. Someone creative came up with the name—the framework the committee will work on is called FREE-AI (Framework for Responsible and Ethical Enablement of Artificial Intelligence) in the Financial Sector. The committee will focus on regulatory and supervisory approaches to AI and identify potential risks. It's great that the banking regulator is focusing on AI because banking is a crucial sector where AI use must be fair and ethical. 

RBI Sets Up 'FREE-AI' Committee To Develop AI Framework

Open source continues to make an impact as IBM reports that more 80% of respondents in a global survey of IT decision makers said that at least a quarter of their AI solutions or platforms are based on open source. Nearly two-thirds of respondents indicated they would increase their AI investments in 2025 and the report said that surveyed companies utilising open-source ecosystems are more likely to be achieving positive RoI than those who are not.

More Companies Turning To Open-Source AI Tools To Unlock ROI: IBM Study

Government To Establish Data And AI Labs In Tier 2 And 3 Cities

See you in 2025!

-Ivor Soans