A recent study by UiPath has shown that 93% of US IT executives are “extremely or very interested” in applying agentic AI to their business, with 45% ready to invest this year. This growing interest shows a huge shift in the world of artificial intelligence, moving beyond the capabilities of current AI systems, such as ChatGPT, which primarily focus on generating content in response to specific prompts. Instead, we are entering the age often referred to as “The Third Wave of AI” i.e. agentic AI, the next generation of AI that can operate more autonomously, making decisions and taking actions with minimal human intervention. This shift has the possibility to overturn various industries and aspects of our daily lives.
Read moreTag Archives: aiml
Is India actually AI ready? Do we have AI for Bharat?
Is Indiaโs AI revolution leaving its own people behind? Weโre building AI for the world, but is India benefiting from its own AI advancements?
Is AI in India empowering or excluding? Who is really benefiting from our AI boom? If AI is the future, why is it still so inaccessible for most Indians?
I know, all these questions might sound very negative, but I really wanted to ask questions precisely, so that we can evaluate better answers for us, us I mean, ๐ช๐จ ๐ฉ๐๐ ๐๐ฃ๐๐๐๐ฃ๐จ.
โ On one side, Indian startups like Uniphore, Fractal Marketing & Analytics, and Innovaccer are making global waves.
โ The Government of India is actively pushing AI initiatives, projects like BHASHINI to tackle language barriers and the IndiaAI Mission to build a national AI ecosystem.
But there is another side of the story that we all shy to speak about:
AI in India is booming as a service provider to the world, but, within the country, its benefits aren’t reaching to its own millions of people.
Why?
Read more
AI playing a big role in ICC (Cricket) Champions Trophy ๐
Cricket has always been about gut feeling, experience, and figuring things out after the game. But now, artificial intelligence (AI) and generative AI (GenAI) are changing the way we receive the real-time updates, i.e. right where it’s happening.
Picture this: a fast bowler charging in, the batsman getting ready, and up in the commentary box, an AI’s already guessing where the ball’s gonna go, checking past data, and changing the odds of winning, all before the ball’s even left the bowler’s hand! And yes, this is not some sci-fi talk, it’s what has started happening already.
AI is constantly reading player biomechanics, identifying fatigue before injuries occur, and even suggesting optimal field placements based on a batsmanโs weaknesses.
And when you think further, even broadcasters are leveraging AI based storytelling, which helps them to personalize highlights based on fan preferences, and even translating commentary into multiple languages in real time. At the same time, social media AI engines are tracking sentiment shifts with every six, wicket, and turning point, giving a real-time pulse on fan reactions across the globe.
So the point is that, AI is saving thousands of research hours that once went into manual video analysis, performance tracking, and match reviews. The result? Faster insights, sharper strategies, and an enhanced viewing experience like we have never seen before.
In this newsletter we will explore the same.
Read more
DeepSeek AI costed me $2675 in 12 days ๐ก
It was a real fight to deploy a large-scale LLM. DeepSeek AI was becoming the star, and so its 67B model appeared exciting to me as I was curious to see if I could make this model work outside of an enterprise-grade setup.
Unlike researchers working at OpenAI, Google, or Meta, I donโt had access to a corporate high-performance computing (HPC) cluster. No racks of NVIDIA H100 GPUs, no enterprise-scale TPU pods, no institutional grants funding for my experiment. But with all these buzz around the low-cost and efficient DeepSeek model, I was very much fascinated.
I immediately saw the potential of running a powerful LLM on my own infrastructure, so that I can research, experiment, and try some custom applications. It was something that I wanted to explore because a lot of my enterprise clients were asking questions for on-premise deployment, which were raising a lot of what-ifs in my mind.
๐๐ก๐ ๐ค๐๐ฒ ๐ช๐ฎ๐๐ฌ๐ญ๐ข๐จ๐ง ๐ฐ๐๐ฌ…
Is it even feasible to deploy such a large model on a cloud platform like AWS or GCP as an individual?
- Can I afford the compute costs without burning through thousands of dollars?
- Will a CPU-only approach work, or is it non-viable?
At first, I thought cloud computing would make this relatively easy. After all, Amazon Web Services (AWS) and Google Cloud Platforms have this on-demand access to high-performance hardware. But…
๐ ๐ฐ๐๐ฌ ๐ฐ๐ซ๐จ๐ง๐ .
Read more
ChatGPT or DeepSeek AI – which one is the right choice for you?
OpenAI ChatGPT and DeepSeek AI is not only learning chatbots, theyโre full-fledged AI assistants that can write, code, problem-solve, and even weigh in on ethical questions. I have been working with them side-by-side since past one month and my findings stunned me in more ways than I can explain. So, drafting my learnings so far with the intent that it might help you to decide which one is the right choice for yourself.
Letโs break it down:
๐ ChatGPT (OpenAI): It is known for its smooth, natural conversations, creativity, and adaptability. I believe, its a go-to for content creation, brainstorming, and engaging dialogues.
๐ DeepSeek: Cool and budget-friendly option that leans on logic, technical accuracy, and efficiency. It is more focused on getting precise answers than creative storytelling.
But here are the parameters I evaluated them on:
๐น Performance: Which one responds faster and with more reliability?
๐น Cost vs. Value: Is DeepSeekโs lower price worth it?
๐น Coding & Content: Who does better at complex programming and writing tasks?
๐น Ethics & AI Safety: How do they handle biases and responsible AI use?
AI cannot be perceived as RPA anymore โ The rise of AI ๐
Artificial Intelligence (AI) was initially confused with RPA (Robotic Process Automation), which focused on automating repetitive tasks. RPA was essentially rule-based, handling processes like data entry, invoice processing, or customer service inquiries. It wasnโt โsmartโ in the way AI is todayโit followed strict, predefined commands.
Back in 2000s, RPA was the closest thing to AI that many companies were implementing. It made work faster but didnโt involve any decision-making or learning. Think about early chatbots or even Microsoft Clippy Assistant. These tools could execute commands but couldn’t “learn” from their interactions.
A comprehensive list of 2024 AI predictions ๐ฝ
The global AI market was already worth more than US$150 billion by the end of 2023. According to one of the reports, the global AI market will reach US$1350 billion by 2030, and this upward journey surely begins from 2024. The idea of this icon ๐ฝ might have set the tone of my predictions for this year. Trust me, we are going to witness unimaginable AI implementations in 2024. The breakthroughs of Generative AI in 2023 has setup a dramatic momentum for 2024, and our expectations have risen to a next level. Everyone is waiting for the “Aliens” to appear this year, I mean not literally, but I guess you understand the sentiments.
Before I start, there are pretty obvious things which are going to happen in 2024, like OpenAI‘s GPT-5 will be launched, Generative AI will become a technology risking most jobs by any tech-disruption, and on contrary setting up stage for people with plethora of opportunity in new job-roles โ like prompting efficiently. And the start of the year will face a dramatic AI startup-stress because of the business models that are too much affected by OpenAI’s release of add-ons.
“It has become appallingly obvious that our technology has exceeded our humanity” โ Albert Einstein
I am very optimistic about this year, but at the same time I am cognizant of the fact that this year is also going to daunt us a lot. And its because we haven’t yet lifted ourselves to the maturity which this technology demands, and I am specifically concerned with the pace of Generative AI’s access to common people in its raw form. So to start with, the table of contents below should clear how I am picturing this for the year 2024.
2023 AI Wrap-up (auto9mous Newsletter)
Its amazing to see how AI adoption happened in the year 2023. With all the Generative AI use cases and new products, last year felt like evolving at an unprecedented speed. Reading through all the stories and posts by AI Influencers, Entrepreneurs, Domain Experts, and literally “Common People“, was so overwhelming. I would call the year 2023, an year of defining the new AI tech-disruption of this decade…
How Responsible a Generative AI Can Be, If We, The Humans, Are Irresponsible?
In the midst of the Fourth Industrial Revolution, as we increasingly integrate Generative AI into our daily lives, we face a critical paradox: Can we expect a machine, inherently void of morals or consciousness, to be responsible if we, the architects and users, sometimes falter in our own responsibilities?
Generative AI systems, like OpenAI’s GPT series or the imaginative Midjourney, have not just demonstrated capabilities to create text or images but have also exemplified the power to inspire, innovate, and occasionally intimidate. Trained on vast troves of data, they’re a mirror, reflecting the collective knowledge, biases, and intentions of humanity.
Before we delve deep, let’s set the context:
Real-world Scenario: In 2020, generative models birthed ‘deepfake‘ technologies, a double-edged sword capable of creating realistic yet entirely synthetic media. While artists found new avenues for creativity, malicious actors found ways to spread misinformation, impacting political landscapes and individual lives.
“A tool is but an extension of one’s hand, an AI is an extension of one’s mind. Both amplify intent; neither possess their own.”
| Generative AI | Typical Usage | Potential Misuse |
|---|---|---|
| GPT-4 | Content Creation, Customer Support | Spreading misinformation |
| MidJourney | Image Generation | Creating misleading imagery |
To visualize the evolution and potential implications of Generative AI, consider this simple flowchart:
This blog will uncover the mechanics of Generative AI, examine the landscape of human responsibilities, and ascertain whether there’s a ceiling to how responsible an AI can truly be. But remember, every tool, even AI, requires judicious and mindful human use. The question isn’t just about what AI can do, but more crucially, what we do with AI.
Continue reading How Responsible a Generative AI Can Be, If We, The Humans, Are Irresponsible?
Leveraging NLP and OCR for Business Card Text Extraction
In the age of digital transformation, where every piece of information is becoming rapidly accessible and organized, business cards remain one of the few tangible pieces of professional information exchange. While their physical form offers a personal touch, extracting information from them in a quick and efficient manner poses a unique challenge. To address this I have thought to write my approach for business card text extraction in the best possible manner.
In the powerful combination of Natural Language Processing (NLP) and Optical Character Recognition (OCR), NLP enables machines to understand and respond to human language. On the other side, OCR technology converts different types of documents, including scanned paper documents, PDF files, or images taken by a digital camera, into editable and searchable data.
In this blog, we will delve into an innovative method that combines the strengths of both NLP and OCR, specifically the renowned Tesseract-OCR tool, to extract and categorize information from business cards. From identifying specific phone numbers such as office, fax, or mobile numbers to precisely extracting detailed address components like city, state, and country, this technique has shown great potential in revolutionizing the way we process business cards. Join us as we unravel the intricacies of this method and explore its future implications.
| Extraction Step | Description | Example |
|---|---|---|
| Optical Character Recognition (OCR) | Conversion of images of typed, handwritten, or printed text into machine-encoded text. | Image of “John Doe, CEO, XYZ Corp.” โ Text: “John Doe, CEO, XYZ Corp.” |
| Extracting and Classifying Phone Numbers | Identifying various phone number types based on prefixes. | Text: “Office: 123-456-7890” โ Classified as “Office Number” |
| Precision Address Extraction | Parsing the text for precise extraction of address details. | Text: “123 Maple St., Springfield, IL 62704” โ Extracted as: Street – “123 Maple St.”, City – “Springfield”, State – “IL”, Zip Code – “62704” |
| Connected Component Analysis | Identifying regions of interest for block processing based on pixel connectivity. | Image with “John” written closely, and “Doe” spaced apart โ Two components: “John” and “Doe” |
| Future Avenues for Business Card Extraction | Potential advancements in the extraction process. | Using Deep Learning for better contextual understanding. |
| Personal Details Extraction | Parsing text for personal names, designations, and organizations. | Text: “Dr. Jane Smith, Cardiologist, HealthCorp” โ Extracted as: Name – “Dr. Jane Smith”, Designation – “Cardiologist”, Organization – “HealthCorp” |
| Phone Numbers Extraction | Retrieving various phone numbers. | Text: “Fax: 987-654-3210” โ Extracted as “Fax Number” |
| Address Extraction | Parsing for the full address. | Text: “456 Elm St., Suite 7A” โ Extracted as “Address” |
| Zip Code Extraction | Isolating and extracting postal codes. | Text: “… Springfield, IL 62704” โ Extracted as “Zip Code – 62704” |
| City Extraction | Identifying and extracting city names. | Text: “… Springfield, IL …” โ Extracted as “City – Springfield” |
| State Extraction | Pinpointing and extracting state names or codes. | Text: “… Springfield, IL …” โ Extracted as “State – IL” |
| Country Extraction | Recognizing and extracting country names. | Text: “… USA” or “… United States of America” โ Extracted as “Country – USA” |