top of page
fevup-logo.png

GROW WITHOUT BORDERS #11

Be the first to know about the latest developments in the e-commerce world! Stay tuned for weekly updated events, breaking news and important developments. Learn about events happening around the world from Fevup Brands Newsletter.


Contents:


  • YouTube is rolling out new Shopping features to help creators promote their products and increase revenue.


  • The new artificial intelligence model Llama 3 has achieved significant success, leaving other open models behind.


  • Trendyol's LLM (Large Language Model) initiative facilitates access to global markets by removing the language barrier in e-export and offers new opportunities to companies.


  • Amazon Music introduced Maestro, a new artificial intelligence technology that will completely change your music experience.


  • VASA-1, Microsoft's new artificial intelligence tool, makes photos talk.


YouTube is rolling out new Shopping features to help creators promote their products and increase revenue.



YouTube on Tuesday announced new Shopping features that allow creators to create shoppable collections, better plan their shoppable videos, and quickly monetize their old content.


The introduction of these new features comes at a time when TikTok Shop is competing with YouTube Shopping and similar competitive platforms. It is stated that TikTok Shop in the USA aims to increase its revenue up to 17.5 billion dollars this year.


YouTube launched "Shopping Collections" to give creators the ability to curate items from their favorite brands for viewers to browse. Creators can choose a variety of products around a particular theme, such as a daily makeup routine or capsule wardrobe. These collections will appear in the creator's product listing, Shop tab, and video descriptions. Content creators will be able to create these collections in the Studio app on their phones. YouTube plans to offer this feature to desktop users in the near future.


The new artificial intelligence model Llama 3 has achieved significant success, leaving other open models behind.



Meta shared a blog post announcing its new large language model (LLM), Llama 3. Llama 3 comes with two different model weights with 8 billion and 70 billion parameters. This model will be available on various platforms such as AWS, Databricks, Google Cloud, Hugging Face, Kaggle, IBM WatsonX, Microsoft Azure, NVIDIA NIM and Snowflake.


Llama 3, which currently only offers text-based responses, is stated by Meta to be a major improvement compared to the previous version. According to the information shared, Llama 3's answers show more variety and have better reasoning abilities. It is also stated that the model understands instructions better and its ability to write code is better than before.


In the post, Meta also shares a table comparing Llama 3 with other models. Accordingly, both size models of the Llama 3 appear to outperform similarly sized models such as Google's Gemma and Gemini, Mistral 7B, and Anthropic's Claude 3.


Trendyol's LLM (Large Language Model) initiative facilitates access to global markets by removing the language barrier in e-export and offers new opportunities to companies.


Trendyol, which brings Turkish products and manufacturers together with international markets, aims to provide services in different languages and facilitate export processes by overcoming the language barrier of sellers with its versatile local language model Trendyol LLM.


Trendyol's Turkish engineers played a leading role in the development process of Trendyol LLM. Currently developed for general use, Trendyol LLM will be offered as open source to the Turkish technology ecosystem and can be customized by companies according to their own needs.


Trendyol's model, which is offered to the Turkish technology ecosystem as open source, has the potential to be used in various fields as it can be customized according to need. The user experiences created by this model are expected to lead to significant developments in many areas, from the development of chatbots to search infrastructures for information extraction. While Trendyol is making plans to customize LLM as a digital assistant for sellers, this innovative model is aimed to be the gateway of Turkish sellers to global markets in the future.


Amazon Music introduced Maestro, a new artificial intelligence technology that will completely change your music experience.


Amazon Music announced an innovation that excites music lovers: the artificial intelligence-supported playlist creation feature called "Maestro"! After Spotify, Amazon Music users will now be able to create playlists that suit their musical tastes with the help of artificial intelligence.


Redefining the music listening experience, Amazon Music offers a unique service to its users with its artificial intelligence feature called Maestro. Competing with Spotify's similar feature, Maestro allows users to create playlists in the style they want with text commands. Moreover, it will be possible to give emotional clues to artificial intelligence by using emoji in commands. This feature gives users wide freedom to personalize the music listening experience.


VASA-1, Microsoft's new artificial intelligence tool, makes photos talk.


Microsoft announced VASA-1, an AI tool that lets you add a voice sample by uploading a still photo and create videos that look and sound like a real person. VASA-1 takes a single portrait-style image and an audio file, combines them, and can create a short video of a talking head with realistic facial expressions, head movements, and even a voice that can sing with the uploaded voice.


Although Microsoft stated that VASA-1 is currently only a research project and therefore not available for general use, it shared interesting videos about the tool. The company explained that the new tool was designed specifically for the purpose of animating virtual characters, and therefore all individuals in the examples are synthetic and created using OpenAI's DALL-E rendering model.


Source:


https://5mid.com

https://webrazzi.com

https://techcrunch.com

Comments


bottom of page