Digital Bhutan

Tag: Artificial intelligence

  • Google Bard directly competes with ChatGPT

    Google Bard directly competes with ChatGPT

    Google has just announced that it is opening up access to its new AI chatbot tool, Bard, which directly competes with ChatGPT. From Tuesday, users can join a waitlist to gain access to Bard’s features, which include outlining and writing essay drafts, planning events like baby showers, and generating meal ideas based on the contents of their fridge.

    In addition to this, Google has also revealed plans to “thoughtfully” add large language models to search “in a deeper way” at a later time, and will start rolling out the tool in the United States and United Kingdom before expanding it to other countries and languages in the future.

    However, this move comes as tech companies like Google, Microsoft, and Facebook compete to develop and deploy AI-powered tools, following the recent viral success of ChatGPT. Google also announced that it will be bringing AI to its productivity tools, including Gmail, Sheets, and Docs.

    But the launch of Bard has not been without controversy, as a demo last month provided an inaccurate response to a question about a telescope. This caused Alphabet, Google’s parent company, to experience a 7.7% drop in share prices, wiping $100 billion off its market value.

    Despite this, Google plans to continue using human feedback to improve its systems and add new “guardrails” to keep interactions helpful and on-topic. Meanwhile, OpenAI has just released GPT-4, the next-generation version of the technology that powers ChatGPT and Microsoft’s new Bing browser, with similar safeguards.

  • DALE: A Deep Learning Architecture for Generating Photos of Bhutanese Architecture

    DALE: A Deep Learning Architecture for Generating Photos of Bhutanese Architecture

    Bhutanese architecture is renowned for its unique blend of traditional and modern design elements. From ancient temples and fortresses to contemporary residential homes, the architecture of Bhutan reflects the country’s rich cultural heritage and spiritual values. However, capturing the beauty of Bhutanese architecture can be challenging, especially for photographers who are limited by access or budget.

    To address these challenges, a deep learning architecture called DALE (Deep Architecture for Learning Bhutanese Elements) has been developed. This architecture uses a combination of convolutional neural networks (CNNs) and generative adversarial networks (GANs) to generate high-quality photos of Bhutanese architecture.

    The DALE architecture consists of two main components: a generator network and a discriminator network. The generator network is responsible for creating new images based on the input data, while the discriminator network evaluates the authenticity of the generated images. The two networks work in tandem to generate realistic images that closely resemble real photos of Bhutanese architecture.

    To train the DALE architecture, a large dataset of real photos of Bhutanese architecture was used. The generator network was trained to produce images that closely resemble the real photos in the dataset, while the discriminator network was trained to distinguish between real and generated images. This process continues until the generator network produces images that are indistinguishable from real photos.

    Once the DALE architecture is trained, it can be used to generate unlimited numbers of photos of Bhutanese architecture. The generated images can be used for various purposes, including architectural visualization, interior design, and even education.

    The DALE architecture offers a new and innovative way of capturing the beauty of Bhutanese architecture. By using deep learning, it generates high-quality photos that closely resemble real photos, providing photographers, architects, and designers with a powerful tool for showcasing Bhutanese architecture.