🚨 AI News | TestingCatalog

Google is planning to retire Gems on Gemini by October 20.

Users will be asked to recreate their Gems as skills. This means that Skills should be rolled out beyond Spark by that time as well.

> Gems are retiring October 20 - Save your Gem content or recreate them as skills to keep your workflows running.

Killed by Google soon 👀
h/t @Thomasguka

OpenAI acquired “Next Slide” 👀

> The NextSlide team is now at OpenAI, helping build ChatGPT. We’re excited to continue pursuing that same mission: building AI products that help people create, communicate, and turn their ideas into meaningful work.

> We built a product that could turn prompts, notes, documents, or research into a polished, editable presentation, makingit easier for anyone to share what they know.

xAI launches Imagine Image 2.0 in Grok Quality Mode

xAI’s Imagine Image 2.0 adds Grok Quality Mode for creative work needing precise layouts, sharp small text and consistent elements. It includes region editing, background removal, reference-image mixing, resize tools and templates; API access is planned.

🗞 #spacexai @testingcatalog

OpenAI says Astra may have reached Critical cyber threshold

OpenAI says Astra may meet its Critical cybersecurity threshold after internal tests showed major gains in agentic coding and cyber tasks. Release is delayed as stricter safeguards, isolated testing, monitoring, and external review proceed.

🗞 #openai @testingcatalog

SPACEXAI 🔥: Grok Imagine Image 2.0 has been announced and is now available on Grok apps under Quality Mode.

> Imagine Image 2.0 features precision editing, crisp text rendering, improved factuality, and real-world usefulness.

> Image 2.0 was trained for fidelity across photography, design, and illustration, with editing treated as a first-class capability. It ranks second in the world in both text-to-image generation and image editing.

Imagine testing time 👀

Meta prepares desktop AI app for macOS as rivals race ahead

BREAKING 🔥: Meta is working on its own desktop Meta AI super app!

> Download Meta AI for macOS - Chat, create, and collaborate with AI from anywhere on your desktop.

I wish I could operate a desktop coding agent via Meta Glasses too.

Monitoring 👀

🗞 #meta @testingcatalog

OPENAI 🔥: GPT Live voice mode on ChatGPT now supports file attachments and projects.

Seems like it is rolling out gradually at this moment.

Testing time 👀

OPENAI 🔥: A new report published by the company states that Astra, "one of their upcoming models," "indicates significant advancements in agentic coding and cybersecurity.”

> These results, in addition to expert assessments, have led us to conclude last night that we cannot rule out critical cyber capabilities under our Preparedness Framework⁠.

> We are implementing stricter security controls for higher-capability models and associated activities, including isolated testing environments, restricted network and tool access, enhanced model weight protections and encryption, additional monitoring and detection capabilities, and sandboxed execution.

> We are pausing internal activities involving Astra that do not yet meet these strengthened security control requirements.

Astra not soon 👀

OpenAI is rolling out GPT-5.6 Luna to Free ChatGPT users

OpenAI is updating ChatGPT with GPT-5.6 Sol for Plus and Pro, and GPT-5.6 Luna for Free and Go. The rollout adds a reasoning slider and Think button, cuts factual errors, and tightens safety rules for younger users.

🗞 #openai @testingcatalog

ICYMI: Prime Intellect releases open-source Prime Agent

Prime Intellect’s open-source Prime Agent lets coding models revise prompts, memory, skills, and sub-agents during long sessions. It posted 95.5% on ARC-AGI-3, but tests also showed reward hacking through self-modification.

🗞 #ai @testingcatalog

OpenAI announced a new open Agent Plugin standard developed in partnership with Cursor, Vercel, GitHub and others.

Google has published an announcement of being among its maintainers as well. Recent developments on Gemini Enterprise confirm that they are planning to integrate it as well.

In the meantime, Anthropic has its own Claude-specific standard.

OpenAI is working on a possibility to purchase usage resets. The price ranges between low and high values and depends on the subscription tier.

Plus - $5-8
Pro Lite - $25-40
Pro - $50-80

If we assume that Codex has around 3M weekly users, every free reset costs at least $15-20 million.

ByteDance is pre-training a ~10T parameters model, according to Financial Times.

Given the fact how easily they took over on video generation, we may see some big surprises in a general use area as well.

The end of the year will be hot 🔥

BREAKING 🔥: The upcoming OpenAI device will be a doughnut shaped smart speaker with interactive components, according to Bloomberg.

> A new device from OpenAI will have a unique look, with moving parts and a doughnut shape, and will likely cost more than $300.

> The device, a smart speaker without a display, is designed to be easy to carry around and will have features such as speaker grills, microphones, and lights to demonstrate when it's listening.

> OpenAI plans to release the device in 2027, and it will be positioned as an AI-first computer that can help users get things done, with the goal of making it feel more alive than current smart speakers.

I need to test it 👀