#ai-model

[ follow ]
#openai

OpenAI takes another step closer to getting AI to think like humans with new 'o1' model

OpenAI's o1 model shows significant progress in reasoning and problem-solving, performing closer to human-like intelligence.
Experts express skepticism about o1's journey towards achieving artificial general intelligence.

OpenAI's Latest AI Can Cost More Than $1,000 Per Query

OpenAI's o3 model showcases impressive AI performance but comes with exorbitant costs, challenging the sustainability of rapid advancements in AI.

This Week in AI: OpenAI's new Strawberry model may be smart, yet sluggish | TechCrunch

OpenAI's Strawberry AI model focuses on accuracy but may struggle due to its slower response times compared to existing generative AI technologies.

Before launching, GPT-4o broke records on chatbot leaderboard under a secret name

OpenAI's groundbreaking GPT-4o AI model reached the top spot on the Chatbot Arena leaderboard, frustrating experts and setting a new record.

OpenAI Threatening to Ban Users for Asking Strawberry About Its Reasoning

OpenAI restricts users from exploring the reasoning of its AI model, 'Strawberry', contradicting its initial open-source vision.

Google Garners Criticism for Demo After Long Awaited 'Gemini' Release

Google's Gemini is a multimodal AI model that can process various types of inputs, set to rival OpenAI's offerings.

OpenAI takes another step closer to getting AI to think like humans with new 'o1' model

OpenAI's o1 model shows significant progress in reasoning and problem-solving, performing closer to human-like intelligence.
Experts express skepticism about o1's journey towards achieving artificial general intelligence.

OpenAI's Latest AI Can Cost More Than $1,000 Per Query

OpenAI's o3 model showcases impressive AI performance but comes with exorbitant costs, challenging the sustainability of rapid advancements in AI.

This Week in AI: OpenAI's new Strawberry model may be smart, yet sluggish | TechCrunch

OpenAI's Strawberry AI model focuses on accuracy but may struggle due to its slower response times compared to existing generative AI technologies.

Before launching, GPT-4o broke records on chatbot leaderboard under a secret name

OpenAI's groundbreaking GPT-4o AI model reached the top spot on the Chatbot Arena leaderboard, frustrating experts and setting a new record.

OpenAI Threatening to Ban Users for Asking Strawberry About Its Reasoning

OpenAI restricts users from exploring the reasoning of its AI model, 'Strawberry', contradicting its initial open-source vision.

Google Garners Criticism for Demo After Long Awaited 'Gemini' Release

Google's Gemini is a multimodal AI model that can process various types of inputs, set to rival OpenAI's offerings.
moreopenai
#google

Google's latest AI Lumiere creates 5-second videos from still images - StartupNews.fyi

Google has introduced Lumiere, an AI model that generates 5-second videos using the Space-Time-U-Net diffusion model.
Lumiere can perform various content creation tasks, including text-to-video generation and video editing.

Google Gemini 2.0 announced with multimodal image and audio output, agentic AI features

Gemini 2.0 represents Google's move towards fully autonomous, goal-oriented AI agents capable of handling complex tasks efficiently.

Google Rebrands Its AI Chatbot as Gemini to Take On ChatGPT

Google is consolidating its generative AI products under the banner of its latest AI model Gemini, which includes the Bard chatbot and Duet AI.
Gemini Ultra, the most powerful version of Google's AI model, is now available in a premium version of Google's chatbot called Gemini Advanced.

Google gets serious about AI-generated video at Google I/O 2024 | TechCrunch

Google introduces Veo, an advanced AI model for video creation, aiming to rival existing models like OpenAI's Sora.

Google expands 'conversational' generative AI tools for advertisers powered by Gemini

Google is expanding the use of its Gemini AI model for search advertisers.
The AI-generated ads still require final human approval before going live.

Google pauses Gemini's image tool for people after anti-woke' backlash

Google paused Gemini AI model after controversy over failure to depict white people.
Gemini's AI image generation has faced criticism for missing the mark on diversity.

Google's latest AI Lumiere creates 5-second videos from still images - StartupNews.fyi

Google has introduced Lumiere, an AI model that generates 5-second videos using the Space-Time-U-Net diffusion model.
Lumiere can perform various content creation tasks, including text-to-video generation and video editing.

Google Gemini 2.0 announced with multimodal image and audio output, agentic AI features

Gemini 2.0 represents Google's move towards fully autonomous, goal-oriented AI agents capable of handling complex tasks efficiently.

Google Rebrands Its AI Chatbot as Gemini to Take On ChatGPT

Google is consolidating its generative AI products under the banner of its latest AI model Gemini, which includes the Bard chatbot and Duet AI.
Gemini Ultra, the most powerful version of Google's AI model, is now available in a premium version of Google's chatbot called Gemini Advanced.

Google gets serious about AI-generated video at Google I/O 2024 | TechCrunch

Google introduces Veo, an advanced AI model for video creation, aiming to rival existing models like OpenAI's Sora.

Google expands 'conversational' generative AI tools for advertisers powered by Gemini

Google is expanding the use of its Gemini AI model for search advertisers.
The AI-generated ads still require final human approval before going live.

Google pauses Gemini's image tool for people after anti-woke' backlash

Google paused Gemini AI model after controversy over failure to depict white people.
Gemini's AI image generation has faced criticism for missing the mark on diversity.
moregoogle
#open-source

Meta Launches New Llama AI Model, Building Towards the Next Stage

Meta launches Llama 3.3, an efficient AI model now most adopted globally, promising innovation through open-sourcing.

Meta Will Soon Launch The Most Powerful AI Model For Free

Meta offering Llama 3.1 AI model for free with 405 billion parameters.

Meta Launches New Llama AI Model, Building Towards the Next Stage

Meta launches Llama 3.3, an efficient AI model now most adopted globally, promising innovation through open-sourcing.

Meta Will Soon Launch The Most Powerful AI Model For Free

Meta offering Llama 3.1 AI model for free with 405 billion parameters.
moreopen-source
#generative-ai

A new Chinese video-generating model appears to be censoring politically sensitive topics | TechCrunch

AI model Kling by Kuaishou censors politically sensitive topics in China.

Apple releases AI image generation tool called MGIE

Apple has released an image editing AI model called MGIE on GitHub
The model can perform simple and complex image editing tasks based on text instructions

Apple's new generative AI model is a curious first step

Apple researchers have released a new AI model capable of making edits to images based on text prompts.
The tool, titled MGIE, offers instruction-based image editing for various aspects.

Samsung unveils Gauss2 AI model with an on-device option

Samsung launched Gauss2, an advanced generative AI model, enhancing performance and efficiency significantly compared to its predecessor.

Microsoft launched a new 'air-gapped' AI model - but none of you can use it

Microsoft created an air-gapped, secure generative AI model for US intelligence agencies.

VERSES' Advances Generative Modeling with Scale-Free Active Inference and Spatial Web Protocol | HackerNoon

New RGM approach by VERSES AI promises faster, less data-intensive, and more efficient AI models compared to traditional methods.

A new Chinese video-generating model appears to be censoring politically sensitive topics | TechCrunch

AI model Kling by Kuaishou censors politically sensitive topics in China.

Apple releases AI image generation tool called MGIE

Apple has released an image editing AI model called MGIE on GitHub
The model can perform simple and complex image editing tasks based on text instructions

Apple's new generative AI model is a curious first step

Apple researchers have released a new AI model capable of making edits to images based on text prompts.
The tool, titled MGIE, offers instruction-based image editing for various aspects.

Samsung unveils Gauss2 AI model with an on-device option

Samsung launched Gauss2, an advanced generative AI model, enhancing performance and efficiency significantly compared to its predecessor.

Microsoft launched a new 'air-gapped' AI model - but none of you can use it

Microsoft created an air-gapped, secure generative AI model for US intelligence agencies.

VERSES' Advances Generative Modeling with Scale-Free Active Inference and Spatial Web Protocol | HackerNoon

New RGM approach by VERSES AI promises faster, less data-intensive, and more efficient AI models compared to traditional methods.
moregenerative-ai

Predictions for the Future of Startups | HackerNoon

The AI Model race equalization, b2b SaaS commoditization, startup focus on real sectors like bio and healthcare, hardware resurgence, platform evolution, distribution significance, audience-centric product development, importance of government APIs, VC funding shift, and the rise of indie makers.

Novel AI Model May Enhance Health Data Interoperability - News Center

AI model FHIR-GPT efficiently transforms EHR data to FHIR resources, advancing health data interoperability and research support.
#meta

Meta's most advanced and powerful LLAMA 3.1 AI opensource model released

Meta released LLAMA 3.1, an opensource AI model, setting a new standard for AI and providing competition to private companies.

Meta releases ChatGPT-like AI site and open-weights Llama 3 model

Meta unveils Llama 3 AI model for text composition, code generation, and chatbots, positioning it against other AI assistants.

Meta's most advanced and powerful LLAMA 3.1 AI opensource model released

Meta released LLAMA 3.1, an opensource AI model, setting a new standard for AI and providing competition to private companies.

Meta releases ChatGPT-like AI site and open-weights Llama 3 model

Meta unveils Llama 3 AI model for text composition, code generation, and chatbots, positioning it against other AI assistants.
moremeta

'Gaslit me so hard' - Coffeezilla renews his assault on Rabbit R1 AI

Coffeezilla exposed Rabbit R1 AI device's lack of genuine AI model, revealing it as a marketing scam.

Meet Falcon 2: UAE's answer to Big Tech's AI

The UAE's TII introduced Falcon 2, a next-gen, open-source, multilingual, multimodal AI model, with vision-to-language capabilities for various industries.

Scientists Claim AI Breakthrough to Generate Boundless Clean Fusion Energy

Stumbling blocks to nuclear fusion include energy generation, reactor materials, impurity control, and fuel containment.
Researchers from Princeton develop AI model to predict and prevent plasma instability in donut-shaped tokamak reactors.

Google's image-generating AI gets an upgrade | TechCrunch

Google announced Imagen 3, an upgraded image generative AI model, with improved accuracy and reduced errors for text prompt translations.

Watch the Google I/O livestream here

Google I/O event offers more than just Pixel 8a release.

Google DeepMind's Latest AI Model Is Poised to Revolutionize Drug Discovery

AlphaFold 3 AI model by Google DeepMind predicts structure of various biological molecules, revolutionizing biological research.

Mystery Deepens as Super Powerful AI Appears, Then Disappears

The mysterious AI model 'gpt2-chatbot' appeared briefly and garnered hype for potentially surpassing GPT-4 capabilities, sparking speculation and comparisons in the AI community.

The Morning After: Senate passes the bill that could ban TikTok

The Senate approved a bill mandating ByteDance to sell TikTok in a bipartisan vote, with President Biden likely to sign it.

Adobe's impressive AI upscaling project makes blurry videos look HD

Adobe researchers developed VideoGigaGAN, an AI model for upscaling blurry videos without introducing artifacts, blending the best of GAN models and traditional upscaling methods.
#phi-3-mini

Microsoft shrinks AI down to pocket size with Phi-3 Mini

Microsoft's Phi-3 Mini AI model focuses on reasoning, rivaling larger models like GPT-3.5 and can run on smartphones offline.

Microsoft's lightweight Phi-3 Mini model can run on smartphones

Microsoft introduced Phi-3 Mini, a light AI model for local devices, offering a cost-effective alternative to cloud-based models for small organizations.

Microsoft shrinks AI down to pocket size with Phi-3 Mini

Microsoft's Phi-3 Mini AI model focuses on reasoning, rivaling larger models like GPT-3.5 and can run on smartphones offline.

Microsoft's lightweight Phi-3 Mini model can run on smartphones

Microsoft introduced Phi-3 Mini, a light AI model for local devices, offering a cost-effective alternative to cloud-based models for small organizations.
morephi-3-mini

AI is helping Amazon send fewer small items in comically large boxes

AI model helps Amazon choose best shipping materials, cuts 2M tons of plastic/cardboard since 2015.
#deepfakes

Dystopian Google AI Tool Turns Still Images into Creepy Talking Videos

Google researchers have developed an AI model, Vlogger, that can turn a single image into a moving, talking avatar.
The Vlogger model aims to enhance online communication, education, and creative processes by creating realistic fake people.

Microsoft's New AI Vasa App Makes Photos Talk and Sing | Entrepreneur

Microsoft introduced AI model VASA-1, transforming single pictures and audio into realistic lip-syncing videos, surpassing competitors in quality.

Dystopian Google AI Tool Turns Still Images into Creepy Talking Videos

Google researchers have developed an AI model, Vlogger, that can turn a single image into a moving, talking avatar.
The Vlogger model aims to enhance online communication, education, and creative processes by creating realistic fake people.

Microsoft's New AI Vasa App Makes Photos Talk and Sing | Entrepreneur

Microsoft introduced AI model VASA-1, transforming single pictures and audio into realistic lip-syncing videos, surpassing competitors in quality.
moredeepfakes

Stability AI is giving more developers access to its next-gen text-to-image generator

Stability AI is making Stable Diffusion 3 available to developers via API and a new content creation platform.

Google says running AI models on phones is a huge RAM hog

Google limited AI model to Pixel 8 Pro due to hardware discrepancies
Pixel 8 to receive Gemini Nano in a limited capacity

X's Grok chatbot will soon get an upgraded model, Grok-1.5 | TechCrunch

Grok-1.5 offers improved reasoning in coding and math tasks over Grok-1
Grok-1.5 has a significantly larger context window for better understanding and generating text
#alphabet

Google's Flagship AI Model Gets a Mighty Fast Upgrade

Alphabet's Gemini Pro 1.5 AI model is more powerful than its predecessor and can handle large amounts of text, video, or audio input at once.
The new model's capabilities allow it to analyze lengthy documents such as a 402-page PDF and answer questions about specific actions in a movie.

Where Will Alphabet Stock Be in 5 Years? | The Motley Fool

Alphabet's AI model is crucial for its advertising revenue
Alphabet's stock is undervalued compared to its peers

Google's Flagship AI Model Gets a Mighty Fast Upgrade

Alphabet's Gemini Pro 1.5 AI model is more powerful than its predecessor and can handle large amounts of text, video, or audio input at once.
The new model's capabilities allow it to analyze lengthy documents such as a 402-page PDF and answer questions about specific actions in a movie.

Where Will Alphabet Stock Be in 5 Years? | The Motley Fool

Alphabet's AI model is crucial for its advertising revenue
Alphabet's stock is undervalued compared to its peers
morealphabet

Scientists Gave AI an "Inner Monologue" and Something Fascinating Happened

AI model called Quiet-STaR teaches itself to reason quietly before providing answers
Model shows work, asks for feedback on correctness, and aims to replicate human inner monologue

Weekly Web Design & Development News: Collective #542

Open-source project simplifying video creation with AI model
ESM and TypeScript rewrite of Google's diff-match-patch
PaneForge for creating resizable panes in Svelte apps

GOODY-2 | The world's most responsible AI model

GOODY-2 AI model prioritizes safety and ethical principles
GOODY-2 outperforms competition by over 70% in safety benchmarks

AI can tell a person's sex with 90 percent accuracy

AI model can detect sex through brain scans with 90% accuracy
Historical inconsistency in differentiating male and female brains based on size

Groq creates fastest generative AI in the world, blazing past ChatGpt and Elon Musk's Grok

Groq's AI model is rapidly gaining attention due to its lightning-fast response speed using LPU architecture.
Groq's hardware outperforms competitors like ChatGPT, achieving speeds up to 500 tokens per second.

Amazon AGI Team Say Their AI Is Showing "Emergent Abilities"

Amazon's new AI model, which was not trained on language abilities, is exhibiting state-of-the-art naturalness at conversational text.
The model was trained on 10,000 hours of speech data and scored highest on the researchers' emergent abilities criteria list.

Cutting-edge AI model improves speech clarity by using human perception

Researchers at The Ohio State University have developed an AI model that improves audio quality by combining subjective ratings of sound quality with a speech enhancement algorithm.
The model outperformed other approaches in minimizing noisy audio and the predicted quality scores were strongly correlated to human judgments.

GOODY-2 | The world's most responsible AI model

GOODY-2 is an AI model built with strict ethical principles
GOODY-2 prioritizes safety over accuracy

Recordings of a one-year-old's life train an AI system to learn words

Babies begin learning language through visual and auditory stimuli
A study has developed an AI model that learns language like a child by observing and connecting dots between what it sees and hears

FashionVerse AI mobile app from Hilfiger Ventures is out now| App Developer Magazine

FashionVerse is now available for download on iOS and Android devices and through the Netflix mobile app.
FashionVerse features a game engine and AI model that create photorealistic avatars and settings to provide an immersive virtual fashion experience.

AI's Latest Challenge: The Math Olympics

AI model can solve geometry problems nearly as well as human gold medalists
AlphaGeometry solved 25 out of 30 Olympiad geometry problems
[ Load more ]