A naukri.com initiative
Data Science News
Analyticsindiamag
1M
420
Image Credit: Analyticsindiamag
What is Nikhil Malhotra’s Dream AI?
- Nikhil Malhotra, the global head of Makers Lab at Tech Mahindra, coined the idea of ‘Dream AI’.
- Dream AI is an architecture combining symbolic AI with deep reinforcement learning, marking a shift from conventional models and addressing fundamental limitations of today’s AI.
- Dream AI builds on a neurosymbolic architecture, drawing from two foundational schools in AI—connectionist (or deep learning) and symbolic (logic and symbols).
- Dream AI aspires to create agents that “dream” by simulating environments and learning with a nuanced understanding of the context, instead of relying on vast datasets alone.
- Reinforcement learning (RL) serves as a bridge between symbolic world models built through simulation engines like the NVIDIA Omniverse and actions guided by deep neural networks.
- Dream AI aims to reduce the training burden on AI systems by using symbolic structures to “dream” or simulate scenarios effectively minimising repetitive data input.
- The integration of symbolic reasoning with deep learning enables a level of adaptability and contextual awareness often missing in autoregressive models.
- Traditional systems are confined to specific tasks, lacking a broader understanding of real-world contexts. Dream AI, however, allows agents to simulate world views, thereby aligning their actions with physical and logical principles.
- By reducing reliance on extensive real-world data, Dream AI makes training both more efficient and scalable.
- Dream AI is a powerful solution for dynamic, complex environments, enabling agents to learn contextually rather than reactively.
Read Full Article
25 Likes
Analyticsindiamag
1M
335
Image Credit: Analyticsindiamag
This Bengaluru Startup is Building a Cheaper Alternative to Salesforce and Zoho
- Tablesprint is an AI-powered low-code development platform aiming to bring enterprise-level applications accessible to all business sizes, designed to compete with established players such as Zoho and Salesforce.
- Traditional low-code platforms often provide limited control over sensitive data. Tablesprint’s platform incorporates data controls, permissions, and a secure back-end infrastructure that integrates with AI, ensuring businesses can leverage their data safely and effectively.
- Tablesprint offers a Salesforce-like experience at a fraction of the cost, positioning it as an alternative to established players like Salesforce and Zoho and stands out as a unique blend of affordability and flexibility.
- Tablesprint’s AI-first SaaS platform allows companies to rapidly build customizable applications for various business functions like HR, sales, operations, and vendor management with the help of both open-source and closed-source models.
- Tablesprint targets the market space by offering a similarly robust platform that’s also simpler to use, much more affordable, and has an optimised user interface (UI) to make processes intuitive and efficient.
- Tablesprint's go-to-market strategy involves partnering with implementation providers, and the company plans to expand its offerings by building a complete AI-based ERP system and form builder.
- Tablesprint is not just an India-specific product; it’s a global solution designed to help businesses worldwide streamline operations, build apps, and generate revenue by offering our platform as a customizable solution.
- Tablesprint’s model and global reach will help the company thrive in the long term, although bandwidth is sometimes an issue due to much interest, pick and choose clientele is essential to ensure each implementation is successful and adds value.
Read Full Article
20 Likes
Analyticsindiamag
1M
205
Image Credit: Analyticsindiamag
Andrew Ng Releases New Short Course on Building Safe & Reliable AI
- DeepLearning.AI has released a new short course called "Safe and Reliable AI via Guardrails" in partnership with Guardrails AI, taught by founder Shreya Rajpal.
- The course focuses on identifying and mitigating common failure modes in LLM-based apps, such as hallucinations and exposure to sensitive information.
- Participants will learn to implement 'guardrails' to prevent these failures and set up guardrails on a pizzeria's chatbot to limit output and detect sensitive information.
- The course only requires basic Python knowledge as a prerequisite and is part of DeepLearning.AI's efforts to provide free courses on safe and efficient deployment of generative AI products.
Read Full Article
12 Likes
Nvidia
1M
350
Image Credit: Nvidia
Indonesia Tech Leaders Team With NVIDIA and Partners to Launch Nation’s AI
- Indonesia’s technology leaders have launched an initiative to bring sovereign AI to the nation’s more than 277 million Indonesian speakers.
- The collaboration is grounded in a broad public-private partnership that reflects the nation’s concept of “gotong royong,” a term describing a spirit of mutual assistance and community collaboration.
- Sahabat-AI is a collection of open-source Indonesian large language models (LLMs) that local industries, government agencies, universities and research centers can use to create generative AI applications.
- Built with NVIDIA NeMo and NVIDIA NIM microservices, the models were launched today at Indonesia AI Day.
- Artificial intelligence will democratize technology. It is the great equalizer,” said Jensen Huang, NVIDIA founder and CEO.
- NVIDIA subsidiary Lintasarta built the high-performance AI cloud within three months, bolstered by the NVIDIA Cloud Partner program.
- IOH is developing applications for industry-specific use cases based on its new AI cloud, Sahabat-AI and the NVIDIA AI Enterprise software platform.
- Hippocratic AI is using the models, the NVIDIA AI platform and IOH’s sovereign AI cloud to develop digital agents that can have humanlike conversations, exhibit empathic qualities, and build rapport and trust with patients across Indonesia.
- Indonesia’s leading on-demand transport services, GoTo, is adopting and enhancing the new Sahabat-AI models to integrate with its AI voice assistant, called Dira.
- Beyond Indonesia, NVIDIA NIM microservices are bolstering sovereign AI models that support local languages in India, Japan, Taiwan and many other countries and regions.
Read Full Article
20 Likes
Medium
1M
330
Image Credit: Medium
How I Used AI to Analyze My Bank Statement: A Game-Changer Over Traditional Excel
- Using Excel for bank statements can be time-consuming and error-prone.
- AI can interpret, categorize, and summarize bank statement data almost instantly.
- Uploading the bank statement to an AI platform removes the need for manual data entry and complex formulas.
- AI organizes transactions, categorizes expenses by vendor, and generates precise totals automatically.
Read Full Article
19 Likes
Analyticsindiamag
1M
375
Image Credit: Analyticsindiamag
AI Just Stopped 5G from Dying
- AITRAS, the world's first AI and 5G telecommunications network, has been launched by NVIDIA and SoftBank Corp.
- The breakthrough solution allows for flexible, cost-efficient and GPU-based AI orchestration, according to chairman of AI-RAN Alliance Jinsung Choi.
- Advanced networks that allow AI and 5G to function correspondingly can unlock newer and higher revenue opportunities for telecom providers, according to Ericsson.
- Satellite internet services like Starlink could work alongside 5G networks, or even compete with them.
- Quantum networking could make 5G less essential if advances in quantum communication lead to a quantum internet.
- AI applications require faster and more reliable networks. 5G may eventually yield to future solutions like 6G, satellite internet, and the outernet.
- 6G, anticipated to support AI applications like autonomous vehicles and smart cities, will require major infrastructure investment and technological breakthroughs.
- Outernet is a global satellite network project aiming to offer free internet access.
- Overall, 5G has marked a significant advancement in wireless communication but future solutions are needed to meet the demands of the AI-powered world.
Read Full Article
22 Likes
VentureBeat
1M
283
Microsoft brings AI to the farm and factory floor, partnering with industry giants
- Microsoft has developed a new suite of specialist AI models created to target specific challenges in agriculture, manufacturing and the financial services industry
- By collaborating with partners like Siemens, Bayer and Rockwell Automation, Microsoft is aiming to bring advanced AI technologies to sectors that have traditionally relied on more traditional methods and tools
- This reflects a broader company strategy that seeks to move beyond general purpose AI and into solutions that can offer immediate operational improvements to industries such as agriculture and manufacturing, under increasing pressure to innovate
- Microsoft’s Phi family of small language models (SLMs) are playing a critical role in this initiative, designed to carry out specific tasks while consuming less computing power
- This may address the limited computing resources available in sectors such as manufacturing, where companies need efficient AI models that can operate effectively on the factory floor
- Early adopters of the systems such as Swire Coca-Cola USA have seen great potential for gains in both efficiency and decision-making
- The initiative also extends into the automotive and finance sectors, with AI models designed to assist car drivers with a voice assistant and to help financial institutions with regulatory compliance
- The new system is set to introduce industry-specific AI models that can be custom-built to address specific business challenges, a strategy not adopted by rivals like AWS and Google Cloud
- Microsoft's strategy could accelerate the adoption of AI in sectors like agriculture and manufacturing, facilitating faster AI adoption in sectors that have been slower to embrace new technologies
- The availability of these AI models through Azure AI Studio and Microsoft Copilot Studio speaks to Microsoft's broader vision of making AI accessible to businesses in every sector, helping to bring AI out of the lab and into the real world
Read Full Article
17 Likes
Medium
1M
352
Image Credit: Medium
Apple Says LLMs Are Really Not That Smart
- Apple has raised concerns about the limitations of Large Language Models (LLMs) and their shortcomings.
- LLMs lack genuine understanding, relying on patterns rather than comprehension.
- LLMs can replicate biases from their training data, leading to inaccuracies.
- LLMs struggle with logical reasoning and maintaining context in longer interactions.
Read Full Article
21 Likes
Medium
1M
343
Image Credit: Medium
Text Areas in zero-true
- Text areas are useful when you want the user to input long form text.
- They provide a larger space for users to type and can receive entire paragraphs of feedback.
- For shorter texts, TextInput is recommended.
- Customizations such as labels and colors can be applied to the text area.
Read Full Article
20 Likes
Medium
1M
251
Image Credit: Medium
Pandas vs. Polars: A Performance Benchmark for Modern Data Analysis
- Polars is a high-performance DataFrame library designed for data manipulation and analysis.
- This article compares Pandas and Polars in terms of processing large amounts of data.
- Polars proves to have faster processing times in various scenarios.
- While Polars offers advantages in performance, Pandas still has superior flexibility and can be used in multiple scenarios.
Read Full Article
15 Likes
Medium
1M
146
Image Credit: Medium
Using Vectors in Audio Filters: Concepts, Applications, and Techniques
- Vectors play a fundamental role in audio filtering by providing a means to represent, process, and analyze audio signals.
- From basic FIR and IIR filters to complex multi-band equalizers and real-time effects, understanding vector operations is essential for designing and implementing effective audio processing algorithms.
- By leveraging vectors, engineers can achieve precise control over audio signal manipulation, leading to high-quality audio experiences.
Read Full Article
8 Likes
Medium
1M
50
Image Credit: Medium
Want To Know What’s Expected Of You? Talk To Your Manager
- Navigating new responsibilities and determining your course of action as a manager requires initiative.
- Scheduling a meeting with your manager is essential to clarify expectations.
- Onboarding procedures can vary from company to company, making it important to figure out as much as possible on your own.
- Ensure that you and your manager are on the same page to define your success as an EM.
- Ask questions about your role, what your manager expects of you, and what specific problems you were hired to solve.
- Understand the performance metrics and KPIs vital to your role.
- Clarify expectations around key areas such as working hours and communication.
- Make it a point to ask directly if there are any responsibilities you need to take on that haven't been discussed yet.
- Understand that understanding manager expectations is an ongoing process that requires regular check-ins.
- Take initiative and exercise your agency in navigating this transition.
Read Full Article
3 Likes
Medium
1M
347
DRIFE: Revolutionizing Urban Mobility with Decentralized Ridesharing
- DRIFE is a blockchain-based, decentralized ridesharing platform that aims to empower drivers and passengers.
- Key features include transparency through blockchain technology, zero commission for drivers, community governance, and incentivized rewards.
- DRIFE benefits drivers with higher income potential, fair governance, and independent operation, while passengers enjoy cost-effective, transparent, and secure rides.
- The platform's native token, $DRF, plays a crucial role in the ecosystem, offering utility and rewards for users.
Read Full Article
20 Likes
Medium
1M
155
Image Credit: Medium
We just got accepted Y Combinator Winter 2025 batch. Here’s how…
- We have been accepted into the Y Combinator Winter 2025 batch.
- We have built PaperAI, an open-source collaborative AI research platform.
- PaperAI helps researchers and scholars annotate, discuss, and get quick summaries on complex points in articles or research papers.
- Our goal is to be the go-to platform for academic collaboration and research enhancement.
Read Full Article
9 Likes
Analyticsindiamag
1M
9
Image Credit: Analyticsindiamag
The CUDA Killer
- Nvidia CEO, Jensen Huang, has said that the company's AI engineers’ goal is to “build once, run everywhere” with CUDA. The company plans to maintain its software indefinitely, even programming languages like C which was cited as an example by Huang. Today, five million developers across around 40,000 companies use CUDA with over 300 code libraries, 600 AI models, and support for 3,700 GPU-accelerated applications, catering to diverse computing needs.
- AMD has launched a competitor to NVIDIA’s CUDA called ROCm 6.2, which Sasank Chilamkurthy, the founder of Johnaic, said has an advantage thanks to its support for PyTorch. ROCm was certainly developed to rival CUDA’s popularity, while AMD has been cultivating an ecosystem with products like its new MI325X accelerators for training and inferencing LLM's.
- Vamsi Bopanna, SVP of AI at AMD, explained that ROCm is designed to connect with ecosystem components and frameworks like PyTorch and model hubs like Hugging Face, and AMD also supports various open-source frameworks like ONXX Runtime and more. Bradley McCredie, corporate vice president at AMD, explained that ROCm eliminates dependence on hardware-specific languages like CUDA through frameworks like Triton, which has enabled AMD GPUs with ROCm to offer a more cost-effective option when compared to NVIDIA GPUs with CUDA.
- However, CUDA is a more mature platform than ROCm. NVIDIA's ecosystem around it is extensive, with a large developer community, extensive documentation, and a broad set of tools for debugging and profiling. While AMD is still developing ROCm to catch up, historically, hardware companies have struggled with building and maintaining strong software ecosystems. Using OpenCL as an example, the user pointed out that although it was supported by hardware companies, it failed to develop into a strong competitor due to inconsistent support and a lack of ecosystem investment.
- At present, CUDA is dominant in GPU programming. 95% of the AI chip market is controlled by NVIDIA, but ROCm offers a compelling alternative for companies requiring an open software ecosystem for deploying high-performance AI models optimized for AMD hardware.
Read Full Article
Like
For uninterrupted reading, download the app