US Edition
Have LLMs Finally Mastered Geolocation? - Bellingcat
- Bellingcat analyzed the geolocation accuracy of large language models from OpenAI, Google, Anthropic, Mistral, and xAI in July 2023 using 500 tests on 25 travel photos.
- This study followed concerns about LLMs hallucinating and aimed to assess their real-world reliability in identifying locations from varied rural and urban images.
- The results showed ChatGPT variations slightly outperformed Google Lens, while Google's Gemini model performed worse, confirming that LLMs are valuable but not flawless geolocation tools.
- The analysis included images from diverse settings such as streets in Beirut, Japanese mountain roads, and the Swiss Plateau and emphasized that all tested models sometimes gave incorrect answers.
- These findings imply that while LLMs will increasingly aid open-source investigations, their limitations require cautious use alongside traditional verification methods.
15 Articles
15 Articles
Have LLMs Finally Mastered Geolocation? - bellingcat
An ambiguous city street, a freshly mown field, and a parked armoured vehicle were among the example photos we chose to challenge Large Language Models (LLMs) from OpenAI, Google, Anthropic, Mistral and xAI to geolocate. Back in July 2023, Bellingcat analysed the geolocation performance of OpenAI and Google’s models. Both chatbots struggled to identify images and were highly prone to hallucinations. However, since then, such models have rapidly…
Under the Radar: NCS NEXT on transformation and how generative AI would reshape the focus of its earlier strategic reset - Under The Radar
Opening new frontiers in information technology and pioneering the future of innovation. That is what we are going to talk about today with our guest from a leading technology services firm in Asia Pacific. Founded in 1981, NCS is a subsidiary of Singtel Group that provides what it calls differentiated and end-to-end services to clients with its capabilities in core areas such as application, infrastructure, engineering and cybersecurity. The f…
RAG and LangChain Basics
Introduction Large Language Models (LLMs) like GPT-4 and Claude are incredibly powerful, but they have a significant limitation: they can only work with information they were trained on. What if you want your AI application to access your company's internal documents, real-time data, or domain-specific knowledge? This is where Retrieval-Augmented Generation (RAG) comes in, and LangChain makes implementing RAG systems remarkably straightforward. …
How Brands Can Win Real Estate in AI-Powered Search
In traditional SEO, ‘real estate’ refers to the layout of a search engine results page. Results that appear near the top of the page are considered prime real estate since they’re what users will see first. The further you move down the page, the less favorable the real estate becomes. As an example, Google’s AI Overviews now occupy the very top of the search results, making them the most desirable real estate: Since AI Overviews now appear fo…
The digital travel sector is undergoing a profound transformation driven by artificial intelligence, and Expedia Group is positioning itself at the forefront of this change. In a context where travelers no longer rely exclusively on traditional search engines but instead interact with conversational tools like ChatGPT and Copilot, the company has redesigned its marketing strategy to adapt to this new consumer logic. Jochen Koedijk, Expedia's Chi…
Coverage Details
Bias Distribution
- 50% of the sources lean Left, 50% of the sources lean Right
To view factuality data please Upgrade to Premium
Ownership
To view ownership data please Upgrade to Vantage
