Press "Enter" to skip to content

Posts tagged as “Artificial Intelligence”

Procreate issue “No generative AI” stance

Procreate, the popular iPad design app, has issue a “No generative AI” policy, the announcement came after most of its competitor are developing their own generative AI or some kind of AI tool for their products/services.

The company said on its AI section their website;

AI is not our future.

Creativity is made, not generated.

Generative AI is ripping the humanity out of things. Built on a foundation of theft, the technology is steering us toward a barren future. We think machine learning is a compelling technology with a lot of merit, but the path generative AI is on is wrong for us.

We’re here for the humans. We’re not chasing a technology that is a moral threat to our greatest jewel: human creativity. In this technological rush, this might make us an exception or seem at risk of being left behind. But we see this road less travelled as the more exciting and fruitful one for our community.

At the same time, Procreate CEO James Cuda published a 40 sec video on the official X/twitter account of @Procreate reinforcing their stance on generative AI.

Here’s the full video below;

Here are some of the reaction that Procreate receives regarding their “NO AI” stance;

Generative AI technology is an amazing technology but this technology is trained using other artists’ content and artwork, to the point that their copyrights are violated and without any compensation.

Primate Labs launches Geekbench AI 1.0 to benchmark AI Performance

Primate Labs Inc. creator of Geekbench, has launch the Geekbench AI, an AI benchmarking tool that measures AI performance. The app measures our devices CPU, GPU, and NPU to determine if it is ready for A.I. and machine learning applications.

Previously called “Geekbench ML,” Geekbench AI 1.0 is the first benchmarking application that will attempt to measure AI performances. Geekbench AI runs ten AI workloads, each with three different data types, giving you a multidimensional picture of on-device AI performance. Using large datasets that mimic real-world AI use cases, both developers and consumers can measure on-device AI performance in just a few minutes with Single Precision, Half Precision, and Quantized scores.

Geekbench AI 1.0 will be available on Android, iOS, Linux, MacOS and Windows. It supports multiple AI Framework like TensorFlow Lite, CoreML, TensorFlow Lite, ONNX and OpenVINO. You can checkout Geekbench AI Workloads documentation to know the tests Geekbench AI uses to measure AI performance.

To use use Geekbench AI 1.0 you can download it from here.

OpenAI launched SearchGPT, an AI-powered Search Engine

OpenAI has announce the launch of SearchGPT, a prototype search engine designed to revolutionize the way we interact with information online. This new tool aims to combine the strengths of OpenAI’s advanced AI models with real-time internet data to provide users with fast, accurate, and contextually relevant answers.

SearchGPT is an AI-powered search engine desgined to leverages the capabilities of OpenAI’s ChatGPT, to deliver timely and precise information. Unlike traditional search engines, SearchGPT is designed to respond to queries in a conversational manner, allowing users to ask follow-up questions and receive answers that build on previous interactions.

Unlike the traditional search form, SearchGPT uses a large textbox asking users “What are you looking for?,” this is similar to what Facebook “What’s on your mind?” and X’s “What’s happening?” asks its users. But instead of the usual plain text and links on the search result, SearchGPT will try to organize the summarizes the result then provide a short description with the attribution links.

You can watch the video below for the demo as published by OpenAI;

We’re testing SearchGPT, a prototype of new search features designed to combine the strength of our AI models with information from the web to give you fast and timely answers with clear and relevant sources. We’re launching to a small group of users and publishers to get feedback. While this prototype is temporary, we plan to integrate the best of these features directly into ChatGPT in the future.

OpenAI has partnered with various news publishers, which include organizations like the Wall Street Journal, The Associated Press, and Vox Media, to ensure that high-quality content is highlighted and easily accessible. SearchGPT is currently being tested by 10,000 test users.

Source: The Verge

Google Translate expands to 110 new languages, including 5 from the Philippines

Google just announced that they are adding 110 new languages to Google Translate, this includes 5 languages from the Philippines namely: Bikol, Hiligaynon, Kapampangan, Pangasinan and Waray.

Google explains that the expansion was possible with the use of artificial intelligence (AI) paired with PaLM 2 large language model.

In addition to the 5 languages from the Philippines, Google translate also added support for the following;

  • Afar which is spoken in Djibouti, Eritrea and Ethiopia.
  • Cantonese, which Google says “one of the most requested languages for Google Translate.”
  • Manx is the Celtic language of the Isle of Man. sands of speakers.
  • NKo a standardized form of the West African Manding languages that unifies many dialects into a common language.
  • Punjabi (Shahmukhi) is the variety of Punjabi written in Perso-Arabic script (Shahmukhi), and is the most spoken language in Pakistan.
  • Tamazight (Amazigh) is a Berber language spoken across North Africa.
  • Tok Pisin is an English-based creole and the lingua franca of Papua New Guinea.

With the additional 110 new languages, Google Translate now supports 243 languages. You can find the full list of newly supported languages on Google Translate here.

New Ray-Bans with Meta’s AI are Coming Soon

Newest Ray Ban will be the next best AI Assistant
A report, mentioned that Meta’s Llama 3 model will be integrated into Ray Ban. This will be an improvement to Meta’s previous collaboration with Ray Ban’s smart glasses.

Ray-Ban Meta Wayfarer

The previous Ray Ban smart glasses featured a 12 MP camera, touch sensors for volume, voice control, and live-streaming capabilities. The smart glasses are currently available in the US, UK and Canada.

The development in the new AI model is thanks to Yann Lecun.

Meta’s New AI Developments
Yann Lecun is Meta’s chief AI scientist. His team is currently working on a video-based learning model for AI. V-JEPA (or Video Joint Embedding Predictive Architecture) is a model that aims to replicate the real-world observational capabilities of humans. The new learning model will “train” AI to have a foundation to give way to reasoning capabilities.

Yann LeCun, Meta AI Chief from meta.com

In their blog, Meta says that V-JEPA is all about helping AI develop a sense of “perception” to help improve AI assistants. AI assistants in smart glasses like the Ray Ban help provide contextual information for users of said products.

Yann Lecun’s interview with Lex Friedman revealed interesting developments regarding the  competitor to the ChatGPT model. In the interview (in video below), Yann Lecun mentioned that the company is always developing models that will be integrated into consumer products.

People will be excited to know that Meta will release two smaller Llama 3 models in the next few weeks this April. Meta launched Llama 2 in July of 2023. This year, Meta plans to release their newest AI model around the same time.

Aside from the release date, an interesting take is that Meta’s Llama 3 will reportedly match ChatGPT 4’s capabilities. The company is in a race with OpenAI as more and more companies join the AI hype including Google and Bing.

There are no official dates yet for the release for the new model. Aside from the AR glasses, Meta plans to integrate their AI models into Whatsapp.

Google announced Gemini, the largest and most capable AI model

Google unveiled its latest and most ambitious artificial intelligence project: Gemini. Touted as the “largest and most capable AI model ever created,” Gemini promises to push the boundaries of what’s possible with machine intelligence, potentially impacting every aspect of our lives.

Gemini’s capabilities are truly impressive. It boasts superior performance on massive multitask language understanding (MMLU) tests, surpassing human experts in areas like math, history, medicine, and ethics. This means Gemini possesses not only vast knowledge but also the ability to reason and solve problems effectively.

Beyond its intellectual prowess, Gemini shines in its multimodal capabilities. It can seamlessly process and understand information from different sources, including text, images, and code. This allows it to perform tasks like explaining complex scientific concepts through interactive simulations or translating languages with nuanced understanding of cultural context.

One of the most exciting aspects of Gemini is its potential to democratize AI. Google plans to offer various sizes and capabilities of the model, making it accessible to a wider range of users and developers. This democratization could lead to an explosion of new AI-powered applications across various industries, transforming fields like education, healthcare, and even creative arts.

Google’s Gemini 1.0, will be available in three different sizes:

  • Gemini Ultra — our largest and most capable model for highly complex tasks.
  • Gemini Pro — our best model for scaling across a wide range of tasks.
  • Gemini Nano — our most efficient model for on-device tasks.

Google acknowledges the ethical concerns surrounding powerful AI and emphasizes its commitment to responsible development. The company has implemented safeguards to ensure Gemini’s safety and prevent misuse.

The launch of Google Gemini marks a significant milestone in the evolution of artificial intelligence. Its capabilities promise to revolutionize the way we interact with technology, solve problems, and understand the world around us. While challenges remain, one thing is clear: the future of AI is bright, and Google Gemini is poised to illuminate the way.

Amazon intros Q, a new generative AI-powered assistant for AWS customers

Amazon just announced Q, a generative AI-powered assistant that aims to “answers to pressing questions, solve problems, generate content, and take actions using the data and expertise found in your company’s information repositories, code, and enterprise systems.”

From their press release;

Amazon Q provides information and advice to employees to streamline tasks, accelerate decision making and problem solving, and help spark creativity and innovation at work. Designed to meet enterprise customers’ stringent requirements, Amazon Q can personalize its interactions to each individual user based on an organization’s existing identities, roles, and permissions. Additionally, Amazon Q never uses business customers’ content to train its underlying models. Amazon Q brings generative AI-powered assistance to customers building on AWS, working internally, and using AWS applications for business intelligence (BI), contact centers, and supply chain management to help organizations of all sizes and across industries use generative AI safely. Amazon Q is available to customers in preview, with Amazon Q in Connect generally available and Amazon Q in AWS Supply Chain coming soon.

Amazon Q will compete with OpenAI’s ChatGPT, Google’s Bard and Microsoft’s Copilot, which is also powered ChatGPT. Amazon Q will be available in 2 Pricing plan; Amazon Q Business for $20/month/user, while the Amazon Q Builder will be $25/month/user.

Source: Amazon

Tesla announces Tesla bot, a humanoid robot right out of science fiction

Tesla just concluded their AI Day and one of the highlights of the event is when Tesla CEO Elon Musk announced that the company is building a Tesla bot, humanoid robot that the company named “Optimus.”

According to their website, Tesla bot is a bi-pedal, humanoid robot capable of performing tasks that are unsafe, repetitive or boring. Tesla bot will be five feet, eight inches tall and will weigh 125 pounds, and have a screen for a face.

Elon Musk added that a prototype will be available “sometime next year” and will take advantage of Tesla’s experience with AI that powers the company’s Autopilot driver assistance software in their products.

You can watch the full video of Tesla’s AI day on the company’s Youtube Channel. You can watch the Tesla bot part of the event in the ebedded video below;