News TodAI

Is GPT-5 really worse than GPT-4o? Ars puts them to the test.

It's OpenAI vs. OpenAI on everything from video game strategy to landing a 737.

US government agency drops Grok after MechaHitler backlash, report says

It appears Groks antisemitic rants stopped it from becoming feds go-to chatbot.

Google releases pint-size Gemma open AI model

The new Gemma model is a fraction of the size of most new models.

Meta backtracks on rules letting chatbots be creepy to kids

Meta drops AI rules letting chatbots generate innuendo and profess love to kids.

Upcoming DeepSeek AI model failed to train using Huaweis chips

Startup encountered persistent technical issues using Huawei Ascend chips.

Sam Altman finally stood up to Elon Musk after years of X trolling

Elon Musk and Sam Altman are beefing. But their relationship is complicated.

Is AI really trying to escape human control and blackmail people?

Opinion: Theatrical testing scenarios explain why AI models produce alarming outputs-and why we fall for it.

Google Gemini will now learn from your chats-unless you tell it not to

Gemini will remember this, so it's time to check your privacy settings.

OpenAI brings back GPT-4o after user revolt

After unpopular GPT-5 launch, OpenAI begins restoring optional access to previous AI models.

OpenAI, cofounder Sam Altman to take on Neuralink with new startup

Sam Altman says we could soon have "high-bandwidth brain-computer interfaces."

Why its a mistake to ask chatbots about their mistakes

The tendency to ask AI bots to explain themselves reveals widespread misconceptions about how they work.

Perplexity offers more than twice its total valuation to buy Chrome from Google

Google may soon be ordered to sell Chrome, and Perplexity is ready.

YouTube backlash begins: Why is AI combing through every single video I watch?

Adult YouTubers defend childish viewing habits in fight to block AI age checks.

Musk threatens to sue Apple so Grok can get top App Store ranking

Musk threatens to sue Apple to get Grok to top spot in App Store rankings.

China tells Alibaba, ByteDance to justify purchases of Nvidia AI chips

Move comes after negotiating an export agreement with the US.

The GPT-5 rollout has been a big mess

OpenAI faces backlash as users complain about broken workflows and losing AI friends.

Reddit blocks Internet Archive to end sneaky AI scraping

The Internet Archive confirmed it's in ongoing discussions with Reddit after block.

GitHub will be folded into Microsoft proper as CEO steps down

Microsoft bought GitHub for $7.5 billion in 2018.

LLMs simulated reasoning abilities are a brittle mirage, researchers find

Chain-of-thought AI "degrades significantly" when asked to generalize beyond training.

Trump strikes wild deal making US firms pay 15% tax on China chip sales

The deal wont resolve national security concerns.

Apple brings OpenAIs GPT-5 to iOS and macOS

It's unclear exactly how GPT-5's new approach to model-switching will work here.

Google Gemini struggles to write code, calls itself a disgrace to my species

Google still trying to fix "annoying infinite looping bug," product manager says.

AI industry horrified to face largest copyright class action ever certified

Copyright class actions could financially ruin AI industry, trade groups say.

ChatGPT users hate GPT-5s overworked secretary energy, miss their GPT-4o buddy

I want my GPT-4o back and Ill do anything to get it.

After using ChatGPT, man swaps his salt for sodium bromide-and suffers psychosis

Literal "hallucinations" were the result.

OpenAI launches GPT-5 free to all ChatGPT users

New model claims fewer confabulations, better coding, and "safe completions" approach.

Heres how deepfake vishing attacks work, and why they can be hard to detect

Why AI-based voice cloning is the next frontier in social-engineering attacks.

US executive branch agencies will use ChatGPT Enterprise for just $1 per agency

Questions linger about ideological bias in models as well as data security.

Google search boss says AI isnt killing search clicks

Liz Reid says Google's data shows AI is generating consistent clicks and better experiences.

Researchers design promptware attack with Google Calendar to turn Gemini evil

The team behind the research has worked with Google to mitigate the attack, but what comes next?

Amazon is still struggling to stem the flood of AI-generated fake books

Meta has released DINOv3, a new AI model for universal image processing that doesn't require labeled data.Trained with self-supervised learning on 1.7 billion images and built with 7 billion parameters, DINOv3 handles a wide range of image tasks and domains with little or no...

Anthropic's Claude Opus 4 and 4.1 models can now end conversations if users repeatedly try to get them to generate harmful or abusive content.The feature kicks in after several refusals and is based on Anthropic's research into thepotential psychological stress experien...

Tencent's Hunyuan-Large-Vision sets a new benchmark as China's leading multimodal model

Zhipu AI's GLM-4.5 is yet another open-source Chinese LLM closing the gap with Western models

OpenAI has updated GPT-5 to sound less formal and more personal after users said the model felt too cold.The model will now use phrases like "good question" or "great start" more often, OpenAI said. Internal tests show no increase in flattery,which had been a ...

OpenAI's o3 model outperforms the newer GPT-5 model on complex, multi-app office tasks

Matrix-Game 2.0 provides an open-source option to Deepmind's Genie 3

Tencent's X-Omni uses open source components to challenge GPT-4o image generation

OpenAI is working on AI systems that can tackle problems for hours or even days at a time.In the company's officialpodcast, Chief Scientist Jakub Pachocki and researcher Szymon Sidor share inside stories about building these long-term thinking models, which are designed to p...

GPT-5''s Rough Takeoff, AI Video Blockbusters, India''s Homegrown LLMs, Synthetic Data Generation

GPT-5''s Rough Takeoff, AI Video Blockbusters, India''s Homegrown LLMs, Synthetic Data Generation

Reinforcement Learning Heats Up, White House Orders Muscular AI Policy, Computer Use Gains Momentum, Fine Control of Fine-Tuning

Reinforcement Learning Heats Up, White House Orders Muscular AI Policy, Computer Use Gains Momentum, Fine Control of Fine-Tuning

DeepSeek Sharpens Its Reasoning: DeepSeek-R1, an affordable rival to OpenAI''s o1

DeepSeek Sharpens Its Reasoning: DeepSeek-R1, an affordable rival to OpenAI''s o1

Tumbling Training Costs, Desktop AI Supercomputer, Tighter AI Export Restrictions, Improved Contrastive Loss

Tumbling Training Costs, Desktop AI Supercomputer, Tighter AI Export Restrictions, Improved Contrastive Loss

Open Agentic LLMs Proliferate, Robot Removes Gallbladders, Reasoning Models Boost Emissions, OpenAI Re-Opens

Open Agentic LLMs Proliferate, Robot Removes Gallbladders, Reasoning Models Boost Emissions, OpenAI Re-Opens

Trump Resets AI Policy, Qwen3''s Agentic Advance, U.S. Chips for China, The Trouble With AI Friends

Trump Resets AI Policy, Qwen3''s Agentic Advance, U.S. Chips for China, The Trouble With AI Friends

Power Moves in AI Coding, Moonshot''s Agentic LLM, How to Comply with EU AI Regs, AI Agents Evolve

Power Moves in AI Coding, Moonshot''s Agentic LLM, How to Comply with EU AI Regs, AI Agents Evolve

Grok Raises Questions, Meta Poaches Talent, California Reframes AI Regulations, Multi-Agent Systems Get Stronger

Grok Raises Questions, Meta Poaches Talent, California Reframes AI Regulations, Multi-Agent Systems Get Stronger

How to Make LLMs Commit Blackmail, Robotic Beehive, Walmart''s AI App Factory, Training Web Agents

How to Make LLMs Commit Blackmail, Robotic Beehive, Walmart''s AI App Factory, Training Web Agents

Amazon''s $100 Billion Bet, Meta''s Sensor-Packed Glasses, Anthropic''s Reason-Free Reasoning, Google''s Extreme Weather Prediction

Amazon''s $100 Billion Bet, Meta''s Sensor-Packed Glasses, Anthropic''s Reason-Free Reasoning, Google''s Extreme Weather Prediction

Judge Rules Training AI on Copyrighted Works Is Fair Use, Agentic Biology Evolves, Meta Befriends Alexandr Wang

Judge Rules Training AI on Copyrighted Works Is Fair Use, Agentic Biology Evolves, Meta Befriends Alexandr Wang

Apple Sharpens Its GenAI Profile, Hollywood Joins Copyright Fight, OpenAI Ups Reasoning Quotient, LLM Rights Historical Wrongs

Apple Sharpens Its GenAI Profile, Hollywood Joins Copyright Fight, OpenAI Ups Reasoning Quotient, LLM Rights Historical Wrongs

FLUX.1 Kontext''s Consistent Characters, Benchmarking Costs Climb, Mary Meeker''s Action-Packed AI Report, Better Video Gen

FLUX.1 Kontext''s Consistent Characters, Benchmarking Costs Climb, Mary Meeker''s Action-Packed AI Report, Better Video Gen

DeepSeek-R1 Refreshed, AI''s Energy Conundrum, Agents Get Phished, Machine Translation in Action

DeepSeek-R1 Refreshed, AI''s Energy Conundrum, Agents Get Phished, Machine Translation in Action

Claude 4 Advances Code Gen, How DeepSeek Built V3 For $5.6m, Google I/O Roundup, O''Reilly Versus OpenAI

Claude 4 Advances Code Gen, How DeepSeek Built V3 For $5.6m, Google I/O Roundup, O''Reilly Versus OpenAI

Codex''s Robot Dev Team, Grok''s Fixation on South Africa, Saudi Arabia''s AI Power Play, 4-Bit Efficiency With 16-Bit Accuracy

Codex''s Robot Dev Team, Grok''s Fixation on South Africa, Saudi Arabia''s AI Power Play, 4-Bit Efficiency With 16-Bit Accuracy

Recipes For Reasoning, Open and Compact Code Generator, Looser AI Regulations, More Factual Output

Recipes For Reasoning, Open and Compact Code Generator, Looser AI Regulations, More Factual Output

ChatGPT Grovels, Qwen3 Takes on DeepSeek-R1, Johnson & Johnson Reveals AI Strategy, Easy Reasoning Hack

ChatGPT Grovels, Qwen3 Takes on DeepSeek-R1, Johnson & Johnson Reveals AI Strategy, Easy Reasoning Hack

OpenAI''s Hit Image Generator, Hot AI Startups, Better Recommendations, Music Generation for Pros

OpenAI''s Hit Image Generator, Hot AI Startups, Better Recommendations, Music Generation for Pros

GPT-5 is supposed to be nicer now

Duolingo CEO says controversial AI memo was misunderstood

AI-powered stuffed animals are coming for your kids

Anthropic says some Claude models can now end harmful or abusive conversations

Sen. Hawley to probe Meta after report finds its AI chatbots flirt with kids

ChatGPTs mobile app has generated $2B to date, earns $2.91 per install

US government is reportedly in discussions to take stake in Intel

Cohere hits a $6.8B valuation as investors AMD, Nvidia, and Salesforce double down

Lovable projects $1B in ARR within next 12 months

Google pushes AI into flight deals as antitrust scrutiny, competition heat up

Inside the Box: Aaron Levie on reinvention at TechCrunch Disrupt 2025

Waymo finally has a music experience worthy of its robotaxi

NeoLogic wants to build more energy-efficient CPUs for AI data centers

Google vet raises $8M for Continua to bring AI agents to group chats

How a once-tiny research lab helped Nvidia become a $4 trillion-dollar company

DeepSeek-V3 New Paper is coming! Unveiling the Secrets of Low-Cost Large Model Training through Hardware-Aware Co-design

DeepSeek Unveils DeepSeek-Prover-V2: Advancing Neural Theorem Proving with Recursive Proof Search and a New Benchmark

Can GRPO be 10x Efficient? Kwai AIs SRPO Suggests Yes with SRPO

DeepSeek Signals Next-Gen R2 Model, Unveils Novel Approach to Scaling Inference with SPCT

Beyond Next-Token Prediction? Metas Novel Architectures Spark Debate on the Future of Large Language Models

Sunday Special: Mind-reading brain chips

Google's new model fits on your phone

ALSO: How to create realistic AI voices for your content

Gemini gets major updates

ALSO: How to turn any document into a video presentation

Perplexity makes bold Chrome bid

ALSO: How to generate music with AI

The real story behind GPT-5

ALSO: How to enable access to all previous ChatGPT models

OpenAI in damage control after GPT-5 release

ALSO: How to build your own mobile app with AI

Sunday Special: A trip to a black hole

Robotics Special: Unitree stuns the industry (again)

GPT-5 is (finally) here

ALSO: How to generate a printable storybook with AI

Gemini launches new AI study tool

ALSO: How to optimize your social media content with a self-hosted n8n

OpenAI, Google, and Anthropic release new models

ALSO: How to build an Interactive app with Claude Artifacts

In Otter news, transcription app accused of illegally recording users voices

Google admits anticompetitive conduct in Australia, agrees to modest fine

Linux is about to lose a feature - over a personality clash

Some users report their Firefox browser is scoffing CPU power

Ransomware crews don't care about your endpoint security - they've already killed it

The plan for Linux after Torvalds has a kernel of truth: There isnt one

Tencent doesnt care if it can buy American GPUs again - it already has all the chips it needs

OpenAI's GPT-5 looks less like AI evolution and more like cost cutting

Your CV is not fit for the 21st century - time to get it up to scratch

Hyundai: Want cyber-secure car locks? That'll be GBP49, please

No more 'Sanity Checks.' Inclusive language guide bans problematic tech terms

UK expands police facial recognition rollout with 10 new vans heading to a town near you

Teaching the model: Designing LLM feedback loops that get smarter over time

This researcher turned OpenAI's open weights model gpt-oss-20b into a non-reasoning 'base' model with less alignment, more freedom

That 'cheap' open-source AI model is actually burning through your compute budget

Gartner: GPT-5 is here, but the infrastructure to support true agentic AI isnt (yet)

Google unveils ultra-small and efficient open source AI model Gemma 3 270M that can run on smartphones

Anthropic takes on OpenAI and Google with new Claude AI features designed for students and developers

Google adds limited chat personalization to Gemini, trails Anthropic and OpenAI in memory features

What happens the day after superintelligence?

Ai2s MolmoAct model thinks in 3D to challenge Nvidia and Google in robotics AI

OpenAI brings GPT-4o back as a default for all paying ChatGPT users, Altman promises plenty of notice if it leaves again

The end of perimeter defense: When your own AI tools become the threat actor

Liquid AI wants to give smartphones small, fast AI that can see with new LFM2-VL model

OpenAI adds new ChatGPT third-party tool connectors to Dropbox, MS Teams as Altman clarifies GPT-5 prioritization

Claude can now process entire software projects in single request, Anthropic says

Salesforces new CoAct-1 agents dont just point and click - they write code to accomplish tasks faster and with greater success rates

TD Securities taps Layer 6 and OpenAI to deliver real-time equity insights to sales and trading teams

Study warns of security risks as OS agents gain control of computers and phones

OpenAI is editing its GPT-5 rollout on the fly - heres whats changing in ChatGPT

AIs promise of opportunity masks a reality of managed displacement

From terabytes to insights: Real-world AI obervability architecture

Anthropic revenue tied to two customers as AI pricing war threatens margins

OpenAI returns old models to ChatGPT as Sam Altman admits bumpy GPT-5 rollout

OpenAIs GPT-5 rollout is not going smoothly

ChatGPT users dismayed as OpenAI pulls popular models GPT-4o, o3 and more - enterprise API remains (for now)

Black Hat 2025: Why your AI tools are becoming the next insider threat

OpenAI launches GPT-5, nano, mini and Pro - not AGI, but capable of generating software-on-demand

New persona vectors from Anthropic let you decode and direct an LLMs personality

The initial reactions to OpenAIs landmark open source gpt-oss models are highly varied and mixed

How a vibe working approach at Genspark tripled ARR growth and supported a barrage of new products and features in just weeks

For regulated industries, AWSs neurosymbolic AI promises safe, explainable agent automation

Nabiha Syed remakes Mozilla Foundation in the era of Trump and AI

Minority Report: Now with more spreadsheets and guesswork

Codeberg beset by AI bots that now bypass Anubis tarpit

Boy riding bubble realizes what he's on, asks for more air

Little LLM on the RAM: Google's Gemma 270M hits the scene

Are you willing to pay $100k a year per developer on AI?

Why the UK public sector still creaks along on COBOL

LLM chatbots trivial to weaponize for data theft, say boffins

Should UK.gov save money by looking for open source alternatives to Microsoft? You decide

Forget Foxconn the iPhone factory. AIs made it a server-slinger first and foremost

Which Agent Causes Task Failures and When?Researchers from PSU and Duke explores automated failure attribution of LLM Multi-Agent Systems

In recent years, LLM Multi-Agent systems have garnered widespread attention for their collaborative approach to solving complex problems. However, its a common scenario for these systems to fail at a task despite a flurry of activity.

Dodgy Huawei chips nearly sunk DeepSeek's next-gen R2 model

DARPAsCylon raiderautonomous fighter jet advances to next phase

Vibe coding platform Anything arrives, our hands-on suggests caution

Amazon's $100B DC spend similar to entire Costa Rica GDP

Doctors get dopey if they rely too much on AI, study suggests

Claude Code's copious coddling confounds cross customers

ByteDance Introduces Astra: A Dual-Model Architecture for Autonomous Robot Navigation

ByteDance introduces Astra, an innovative dual-model architecture revolutionizing robot navigation in complex indoor environments.

MIT Researchers Unveil SEAL: A New Step Towards Self-Improving AI

MIT introduces SEAL, a framework enabling large language models to self-edit and update their weights via reinforcement learning.

Researchers from PSU and Duke introduce Multi-Agent Systems Automated Failure Attribution

Automated failure attribution is a crucial component in the development lifecycle of Multi-Agent systems. It has the potential to transform the challenge of identifying what went wrong and who is to blame from a perplexing mystery into a quantifiable and analyzable problem

Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models

By combining State-Space Models (SSMs) for efficient long-range dependency modeling with dense local attention for coherence, and using training strategies like diffusion forcing and frame local attention, researchers from Adobe Research successfully overcome the long-standing ch...

Zhipu.AIs Open-Source Power Play: Blazing-Fast GLM Models & Global Expansion Ahead of Potential IPO

Zhipu.AI open-sources faster GLM models (8x speedup), launches Z.ai, aiming for global expansion, potentially ahead of IPO.

AI Video Generation Race Shifts from Capability to Profitability, Challenging Soras Dominance

The AI video generation landscape is shifting from capability to profitability, challenging OpenAI Soras dominance. Competitors are surpassing Sora in quality and efficiency, with users preferring alternatives. The focus is now on improvements like precise control and style custo...

Nvidia Intensifies Robot Push with New Humanoid Platform as Industry Giants Eye Lucrative Future

Nvidia will launch Jetson Thor for humanoid robots in H1 2025, entering a growing market where Google is also active. The robotics sector is projected for substantial growth. Nvidia offers integrated hardware and software solutions. Simultaneously, Chinas rapidly developing domes...

Automating Artificial Life Discovery: The Power of Foundation Models

A research team introduces Automated Search for Artificial Life (ASAL). This novel framework leverages vision-language FMs to automate and enhance the discovery process in ALife research.

Llama 3 Meets MoE: Pioneering Low-Cost High-Performance AI

Researchers from the University of Texas at Austin and NVIDIA proposes upcycling approach, an innovative training recipe enables the development of an 8-Expert Top-2 MoE model using Llama 3-8B with less than 1% of the compute typically required for pre-training.

DeepMinds JetFormer: Unified Multimodal Models Without Modelling Constraints

A DeepMind research team introduces JetFormer, a Transformer designed to directly model raw data. This model maximizes the likelihood of raw data without depending on any pre-trained components, and is capable of both understanding and generating text and images seamlessly.

NVIDIAs nGPT: Revolutionizing Transformers with Hypersphere Representation

An NVIDIA research team proposes the normalized Transformer, which consolidates key findings in Transformer research under a unified framework, offering faster learning and reduced training steps-by factors ranging from 4 to 20 depending on sequence length.

From Token to Conceptual: Meta introduces Large Concept Models in Multilingual AI

A research team at Meta introduces the Large Concept Model (LCM), a novel architecture that processes input at a higher semantic level. This shift allows the LCM to achieve remarkable zero-shot generalization across languages, outperforming existing LLMs of comparable size.

NVIDIAs Hybrid: Combining Attention and State Space Models for Breakthrough Performance of Small Language Models

An NVIDIA research team proposes Hymba, a family of small language models that blend transformer attention with state space models, which outperforms the Llama-3.2-3B model with a 1.32% higher average accuracy, while reducing cache size by 11.67x and increasing throughput by 3.49...

From Response to Query: The Power of Reverse Thinking in Language Models

In a new paper Time-Reversal Provides Unsupervised Feedback to LLMs, a research team from Google DeepMind and Indian Institute of Science proposes Time Reversed Language Models (TRLMs), a framework that allows LLMs to reason in reverse-scoring and generating content in a manner o...

Yann LeCun Teams New Research: Revolutionizing Visual Navigation with Navigation World Models

In a new paper Navigation World Models, a research team from Meta, New York University and Berkeley AI Research proposes a Navigation World Model (NWM), a controllable video generation model that enables agents to simulate potential navigation plans and assess their feasibility b...

The Future of Vision AI: How Apples AIMV2 Leverages Images and Text to Lead the Pack

An Apple research team introduces AIMV2, a family of vision encoders that is designed to predict both image patches and text tokens within a unified sequence. This combined objective enables the model to excel in a range of tasks, such as image recognition, visual grounding, and ...

Redefining Music AI: The Power of Sonys SoniDo as a Versatile Foundation Model

In a new paper Music Foundation Model as Generic Booster for Music Downstream Tasks, a Sony research team presents SoniDo, a groundbreaking music foundation model that offers robust framework for improving the effectiveness and accessibility of music processing.

DeepMinds Socratic Learning with Language Games: The Path to Self-Improving Superintelligence

Researchers from Google DeepMind introduce the concept of Socratic learning. This refers to a form of recursive self-improvement in artificial intelligence that significantly enhances performance beyond the initial data or knowledge available to the system, as well as a practical...