Long-form thoughts on AI, systems, and the shifting shape of work.
I also publish on Medium. For the cross-posted list, visit medium.com/@prdeepak.babu.
How a design principle from 1970s operating systems is quietly shaping the future of autonomous AI Introduction In 1971, Ken Thompson and Dennis Ritchie faced a problem that sounds oddly familiar tod…
ReadIntroduction I’ve been working in and around AI research and applied machine learning for close to two decades now. I’ve lived through multiple hype cycles, winter cycles, algorithmic plateaus, infra…
ReadInsights, pitfalls, and architectural truths from real-world deployments The first time you demo an agent internally, it feels magical. It answers questions, writes decent text, sometimes even calls…
ReadImagine you deploy a chatbot assistant at your company. It can answer FAQs, lookup policies, book meetings — all sorts of tasks. Early on, it seems to handle single questions well. But when employees…
Read“In science, if you know what you’re doing, you shouldn’t be doing it. In engineering, if you don’t know what you’re doing, you shouldn’t be doing it.” — Richard Hamming This quote resonates deeply w…
Readcross-posting from linkedin Deepak Babu P R on LinkedIn: #coding #ai #hci #reasoning #windsurf #cursor A͟I͟ ͟f͟o͟r͟ ͟c͟o͟d͟i͟n͟g͟ ͟t͟a͟s͟k͟s͟ One of the recurring topics in my coffee chats with scien…
ReadIntroduction: A ChatGPT Moment, Reimagined The release of DeepSeek this week has sparked a familiar buzz — a ChatGPT moment for 2025. Users outside the AI/tech bubble are once again witnessing the ma…
ReadInformation systems have evolved from simple keyword-based search engines to sophisticated natural language-based question-answering (QA) systems. I distinctly remember my early days in 2012, buildin…
ReadAs an AI scientist who has been immersed in the field for many years, I’ve witnessed firsthand the grassroots innovations that have propelled artificial intelligence forward — be it large language mo…
ReadM achine learning (ML) and artificial intelligence (AI) are fundamentally about learning patterns and models from data or a collection of experiences. These experiences can be thought of as individua…
ReadI had a fantastic time at the Databricks Data + AI Summit (DAIS) 2024, connecting with over 2K+ participants from diverse backgrounds, including CXOs and ICs (engineers, scientists, and product manag…
ReadI might be biased, but it seems that Siri’s ability to understand and execute commands has recently taken a significant leap forward. This step-function improvement in Siri’s responses might be due t…
ReadWelcome to “Thinking-Out-Loud,” a series where I delve into intriguing research papers and share my spontaneous reflections, insights, and personal experiences. Each post will explore a different pap…
ReadIn recent years, the field of artificial intelligence has experienced unprecedented growth. Two driving forces underpin this advancement: the availability of colossal datasets, exemplified by the Lla…
ReadSample Efficiency (Data) In the realm of machine learning, “sample efficiency” is a term that defines a model’s ability to learn effectively from a limited number of examples. High sample efficiency…
ReadMultimodal models are creating a synergy between previously separate research areas such as language, vision, and speech. These models use universal architectures that treat each modality as a distin…
ReadWe have been discussing audio and language models in last few posts ( Multimodal audio-language LMM and Qwen-Audio FM ). In continuation to these, here we discuss another important foundation model f…
ReadAs a follow-up to the last blog on large multimodal audio models (LMM), we’re here to explore an open-source large LMM audio model. While dozens of image and text MM models exist, those involving aud…
ReadRecent advancements in Large Language Models (LLMs) have demonstrated a significant capability for reasoning and planning, a trait predominantly seen in models with over 100 billion parameters (consi…
ReadIn the age of foundational models that are based on deep learning architectures like transformer models, we can process large amounts of data to learn complex representations of data, forming world k…
ReadMechanistic Interpretability is a field of study that concerns study of neural networks (more generally ML models) with an intent to understand and explain inner workings of a machine learned model.…
ReadWe often hear reward hacking in the context of LLMs generating content that mimics style over substance. so what exactly does this mean ? why does it occur ? Before we dive into concept of reward hac…
ReadAs we transitioned from the traditional AI methods of the 2010s to deep learning-based approaches in the 2020s, there’s been a marked shift towards leveraging large-scale datasets. These datasets are…
ReadA Comprehensive Look at ‘The Cold Start Problem’ Having been a part of several unicorn startups in India, including Mu Sigma, InMobi, and Snapdeal, I’ve had the opportunity to contribute to and obser…
ReadHaving been a part of several unicorn startups in India, including Mu Sigma, InMobi, and Snapdeal, I’ve had the opportunity to contribute to and observe the unique challenges and dynamics of their gr…
ReadI so wish research community stressed more on this. (negative results) I think some conferences have started to have a forum dedicated to these, but these are not yet mainstream.
ReadI would emphasize, taking an independent shot at the problem first and then refer literature and third, see if existing methods can be improved fusing ideas from of your own from (i) or those resulti…
ReadLLMs are fast-evolving and we have a new model every week, showing up on leaderboards[1] beating previous SoTA model on multiple NLP benchmarks. There are multiple architectures with nuances to train…
ReadEmergent abilities refer to the capabilities that arise spontaneously from the complex interactions of simpler components. They are properties that can’t be predicted solely based on the individual p…
ReadThe growing capabilities of Large Language Models (LLMs) have led to an increasing demand for their integration into production systems. However, implementing LLMs with low-latency requirements like…
ReadLarge Language Models (LLMs) have caused a paradigm shift in the way NLP is traditionally done i.e anything to do with text. I have been working in this domain for 10+ years having seen n-gram ways o…
ReadWe have all been in situations where we have felt a lack of clarity and understanding on a complex project, metrics defined for measurement disagree with customer anecdotes, deploying a solution that…
ReadReinforcement Learning (RL) is a learning paradigm different from traditional machine learning (supervised and unsupervised). The learning problem considered here mimics humans learning from interact…
Readprdeepakbabu/Python Python - All *.py scripts github.com Deep learning neural networks have shown promising results in problems related to vision, speech and text with varying degrees of success. I h…
ReadMachine Learning – Linear Regression using gradient descent vs. neural network Machine learning or Supervised Learning broadly encompasses two classes of problems – regression and classification. Reg…
ReadInformation Retrieval : Spell Correction in Search Engines – Algorithms & Strategies Spell Checking & Correction is an important component of any information retrieval(aka document retrieval) system…
ReadIn this blog post, I am going to discuss about mobile advertising networks, ecosystem & industry trends. This understanding would serve as base for the future posts I plan to write around data scienc…
ReadWe talk about big data quite often these days, wanted to put some fundas about basics around data. Do you know the singular form of data ? How data differs from information vs. knowledge? How insight…
ReadI have spent most of my life into data and its applications to problems. Now, when i look at some patterns in algorithms we use in analyzing data, one thing that emerges is increased use of meta-algo…
ReadWe are in the era of big data, with newer sources of data emerging at an exponential rate involving sensor data, EHR, social network/media data & machine generated data. In this blog post, I will be…
ReadClustering is an unsupervised classification (learning) technique, where the objective is to maximize inter-cluster distance while minimizing the intra-cluster distance. By unsupervised, we mean clus…
ReadHadoop is an open source framework for writing and running distributed application that process huge amounts of data ( more famously called Big Data). The name Hadoop is not an acronym; it’s a made-u…
ReadIn the recent past, a variety of new social media sites have emerged – location Based Services( like foursquare, gowala), Group Deals( like groupon), microblogging( like twitter, fb). These social me…
ReadThis blog post is about the analysis of implementation of helmet rule in various Indian states and the effect it had on bringing down accidental deaths due to 2 wheelers. Here we are specifically foc…
ReadThis blog post is about comparison of amazon.com and linkedin.com in terms of similarities across dimensions of analytic maturity & use of data shared by their customers. As Thomas Davenport mentions…
ReadTurning raw data into insights often involves integrating data from multiple disparate sources (not just limited structured one), analyzing the data, visualizing it and socializing the results/insigh…
ReadIn this blog post, i talk about 3 scenarios where there had been highly valuable insights derived, yet remaining simple. 1. Customers shopped online returned via stores Randy Lea, VP product & servic…
ReadIn my previous post , i had discussed about Association rule mining in some detail. Here i have shown the implementation of the concept using open source tool R using the package arules. Market Baske…
ReadFor the last 6 months, i have been closely following trends in information management. Below are few of my observations. Data source explosion: Business Problems are gaining complexity day by day, he…
ReadDo you find analytics/data mining a difficult topic to understand and learn? To a certain extent true if you were to use books as the source. Friends, i found these two very valuable and high quality…
ReadAssociation Rule Mining [ Implementation using R here] Association Rule mining is one of the classical DM technique. Association Rule mining is a very powerful technique of analysing / finding patter…
ReadHere is the continuation to the article i had posted few days back < here >. I am back with some interesting info on recent advancements in the area of analytics. Before going on to the details, wann…
ReadInformatica recently unvieled its latest release informatica 9 with some major enhancements as compared to its earlier versions, believed to revolutionize data integration & ETL market. I know lot of…
ReadVisualization is considered to be one of the valuable tools in data mining. Visual analysis helps to understand data with minimum effort. Using graphs/ 3D plots to visualize data is more effective wa…
ReadIn my today’s post i will be talking about amazon.com, a leading e-commerce company. Amazon.com has huge amounts of data about its customer base, products and customer purchase behaviour. To boost up…
ReadRecently i had attended a conference on “Future of Predictive Analytics” here at Bangalore. Here is the summarized version of the topics covered; Data mining is an automated process of discovering hi…
Read