a computer chip with the letter a on top of it

Week of August 18-22, 2025

Stay updated on Aug 18–22, 2025: GPT-5 rollout stirs debate, Grok 4 becomes free (with limits), Claude gains memory, Meta licenses Midjourney visuals, and OpenAI eyes GPT-6.
Reading Time: 2 minutes

Aviso de Tradução: Este artigo foi automaticamente traduzido do inglês para Português com recurso a Inteligência Artificial (Microsoft AI Translation). Embora tenha feito o possível para garantir que o texto é traduzido com precisão, algumas imprecisões podem acontecer. Por favor, consulte a versão original em inglês em caso de dúvida.

xAI Makes Grok 4 Free Worldwide

Elon Musk’s xAI has made its flagship model, Grok 4, freely available to all users globally. This strategic move aims to boost accessibility and challenge OpenAI’s market dominance, signaling stronger competition and faster adoption for advanced AI tooling

OpenAI Releases GPT‑5—With Rocky Reception

OpenAI officially unveiled GPT‑5 on August 7, praising its leaps in reasoning, multimodal understanding, and coding proficiency.

However, the rollout was controversial—many users criticized its colder tone and sudden replacement of GPT‑4o, prompting OpenAI to reinstate the previous model for some subscribers.

Anthropic Adds On-Demand Memory to Claude

Anthropic has introduced a new on-demand conversational memory feature for Claude, enabling the chatbot to recall and summarize past interactions when explicitly instructed—currently available to Max, Team, and Enterprise users. This change enhances continuity while preserving user privacy.

Meta Partners with Midjourney for AI Visuals

Meta has struck a deal with AI image/video pioneer Midjourney to license its “aesthetic technology” for future Meta products and models. The partnership signals a strategic shift from Meta’s in-house AI development as it aims to catch up in generative media capabilities.

OpenAI Teases GPT‑6

Just days after GPT‑5’s release, Sam Altman revealed that GPT‑6 is already in development and may arrive much sooner than previous models. Early hints suggest deeper personalization, extended memory features, and faster iteration cycles, though safety and ethics remain key concerns.

a computer chip with the letter a on top of it

Table of Contents

Post Tags

Support this site

Did you enjoy this content? Want to buy me a coffee?

Related posts

man ice skiing on hill
Weekly News
nunobreis@gmail.com

AI Weekly News: Week of January 5–9, 2026

AI Weekly News (Jan 5–9, 2026): Big Tech doubles down on AI infrastructure, agentic systems move into deployment, regulators signal tougher enforcement, regional AI models gain attention, and workforce transformation accelerates.

Read More »
focused photo of a snow flake
Weekly News
nunobreis@gmail.com

Major AI News Highlights: Week of December 8–12, 2025

Weekly AI News (Dec 8–12, 2025): IBM moves to acquire Confluent for $11 B to boost AI infrastructure; Time names AI leaders ‘Person of the Year’; OpenAI warns of high cybersecurity risks in future models; and Google adds source attribution to AI Mode search.

Read More »

Stay ahead of the AI Curve - With Purpose!

I share insights on strategy, UX, and ethical innovation for product-minded leaders navigating the AI era

No spam, just sharp thinking here and there

Level up your thinking on AI, Product & Ethics

Subscribe to my monthly insights on AI strategy, product innovation and responsible digital transformation

No hype. No jargon. Just thoughtful, real-world reflections - built for digital leaders and curious minds.

Ocasionally, I’ll share practical frameworks and tools you can apply right away.