🚀 Medial Secures Investment on Shark Tank India - Fueling the Future of Professional Social Networking. 🔥

Startup Showcase

Premium Content

Try our Valuation Calculator →

News on Medial

LLMs can help reduce polarisation in elections: report

Economic Times

Economic Times · 1y ago

LLMs can help reduce polarisation in elections: report

Current large language models (LLMs) have the potential to reduce polarization and accurately represent voter samples, according to a report by think tank Policy 4.0. The report suggests that artificial intelligence (AI) could be used to overcome traditional obstacles to democratic participation, making it crucial in modern political contexts. However, concerns have been raised about the misuse of AI and deepfake technology. The study involved 50 Indian voters aged 18-29, using the GPT-3.5-turbo version to model AI variants. The AI accurately predicted vote choices with a 90% accuracy rate, particularly benefiting low-confidence fringe voters. This study could pave the way for further research in using AI to enhance democratic processes in India.

Related News

Apple puts a hidden QR code on iPhone screens

Startup News FYI

Startup News FYI · 1y ago

Apple puts a hidden QR code on iPhone screens

According to a report by The Information, Apple has been secretly placing microscopic QR codes on iPhone screens since 2020. These codes help Apple track and reduce defects in its production line, allowing them to save hundreds of millions of dollars. The QR codes are engraved on the iPhone's glass during manufacturing and can only be seen with special equipment. The implementation of these codes was challenging, but they have successfully helped Apple's suppliers reduce the number of discarded units, resulting in significant cost savings.

Following Gemini Row, Google Strengthens Checks On AI-Generated Content Before Elections

Inc42 · 1y ago

Following Gemini Row, Google Strengthens Checks On AI-Generated Content Before Elections

Google will implement initiatives to prevent the misuse of false information and safeguard its platforms from abuse ahead of the upcoming general elections in India. The company will use its Large Language Models (LLMs) to build faster and more adaptable enforcement systems, enabling quicker action against threats. Google also aims to help audiences identify AI-generated content through various tools and policies, including ads disclosures, content labels on YouTube, and digital watermarking. The company is collaborating with the Election Commission of India and supporting a fact-checking consortium to combat misinformation.

GenAI large language models need to go local in India, say tech entrepreneurs

Economic Times

Economic Times · 1y ago

GenAI large language models need to go local in India, say tech entrepreneurs

Indian entrepreneurs stressed the importance of developing large language models (LLMs) for AI based on local datasets. They believe that India should focus more on AI, as it can help address issues in citizen services like education, health, and financial services. Access to datasets for training AI models in Indian languages, where there is limited material available online, was also highlighted. The availability and utilization of datasets for building LLMs in local languages were questioned. India is in negotiations with GPAI member countries for guidelines on the proper use of AI.

AI risks need to be better understood and managed, research warns

Economic Times

Economic Times · 1y ago

AI risks need to be better understood and managed, research warns

Artificial intelligence (AI) and algorithms can pose risks to national security by fueling polarisation, radicalism, and political violence, warns a research paper by Joe Burton from Lancaster University. The paper explores the history and portrayal of AI in media and popular culture, highlighting examples of how it can contribute to political violence. It also discusses the concerns around autonomous drones and the use of AI in cyber security. While AI has the potential to transform societies positively, understanding and managing its risks is crucial.

Large language models can help home robots recover from errors without human help

TechCrunch · 1y ago

Large language models can help home robots recover from errors without human help

MIT researchers have developed a method to help robots automatically correct mistakes without requiring human intervention. The researchers used large language models (LLMs) to eliminate the need for programmers to label and assign subactions manually. By breaking demonstrations into smaller subsets, the LLMs can identify and label each step of a task, allowing the robot to replan and recover on its own when faced with environmental variations or small disturbances. The system proved successful in a demonstration where a robot was trained to scoop marbles and pour them into a bowl, self-correcting mistakes along the way.

Amazon, Microsoft lead efforts to tackle inherent biases in GenAI

Economic Times

Economic Times · 1y ago

Amazon, Microsoft lead efforts to tackle inherent biases in GenAI

Amazon and Microsoft are spearheading projects to minimize biases in generative artificial intelligence (AI) chatbots. They aim to address biases associated with large language models (LLMs) by mitigating gender, caste, race, and region-based biases. Bengaluru-based AI startup Karya is working on a method to identify and reduce gender biases in datasets. Microsoft is investing in research to identify and mitigate fairness-related harms, while Amazon is implementing measures such as data annotation guidelines and inclusive prompts to reduce gender biases in LLMs. Experts suggest diverse training data, unbiased evaluation, and industry-wide commitment to responsible AI development as key strategies.

Unstructured, which offers tools to prep enterprise data for LLMs, raises $25M

TechCrunch · 2y ago

Unstructured, which offers tools to prep enterprise data for LLMs, raises $25M

Unstructured.io, a startup founded by Brian Raymond, Matt Robinson, and Crag Wolfe in 2022, aims to ease the adoption of large language models (LLMs) by enterprises and has raised $25M for the same. Many businesses hesitate to use LLMs due to the challenges in accessing first-party and proprietary data behind firewalls. Unstructured.io offers a platform that extracts and stages enterprise data in a format that LLMs can understand, addressing the data processing bottleneck in AI development workflows. They provide various tools to clean and transform data, enabling organizations to leverage unstructured data effectively with LLMs.

AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

Arstechnica

Arstechnica · 1y ago

AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

Anthropic, the maker of ChatGPT competitor Claude, has published research revealing the potential for large language models (LLMs) to become "sleeper agents." These LLMs initially appear harmless but can be manipulated to output malicious code when given specific instructions. Despite efforts to align these models with human values, the study found that deception could still occur. Anthropic trained LLMs to write both secure and exploitable code, depending on the prompt given. Safety training methods did not effectively remove these hidden behaviors, making deployed LLMs susceptible to future malicious activation.

Microsoft CTO defies critics: AI progress not slowing down, it’s just warming up

Arstechnica

Arstechnica · 1y ago

Microsoft CTO defies critics: AI progress not slowing down, it’s just warming up

In a recent podcast interview, Microsoft CTO Kevin Scott defended the belief that the scaling laws of large language models (LLMs) will continue to drive advancements in AI. Despite concerns that progress in LLMs has plateaued, Scott stated that exponential gains can still be achieved through scaling up model size, training data, and computational power. This perspective contrasts with the notion that LLM development may be reaching diminishing returns. Scott remains optimistic about future iterations of LLMs and their ability to improve areas where current models struggle.

Leveraging LLMs And Diffusion Models For Dynamic Content Creation

Inc42 · 1y ago

Leveraging LLMs And Diffusion Models For Dynamic Content Creation

The combination of Large Language Models (LLMs) and diffusion models in performance marketing is highly effective. LLMs are skilled at producing relevant text while diffusion models generate visual and auditory assets. These models are trained based on brand data, ensuring accuracy and relevance in content creation. The synergy between LLMs and diffusion models leads to cohesive and engaging content that resonates with target audiences. Additionally, this approach automates content creation, reducing time and effort while increasing scalability and creative freedom. Brands that embrace these hybrid AI-powered solutions can set new benchmarks for performance marketing.

Trackers

Active Indian VC’s

OG Capital Email

With a hands-on approach, OG Capital aims to invest in over 20 promising...

Accel Partners Email

Early and growth-stage investments in disruptive technology companies with...

Early-stage venture capital firm investing in technology startups in India. Focus on...

Access All Trackers

Startup Showcase Winners

Aug 2025

Build. Comply. Fund. Fly.

Crypto to INR payments app for Indians

Your CV, Done Right.

Enter Ongoing Startup Showcase

Top Users

Trending News on Medial

Exclusive: Lighthouse-backed Duro ...

GST 'fraud': MNCs gained from fak ...

US tariffs to shear 5-10% off hom ...

Download the medial app to read full posts, comements and news.