Advertise here | 6-min Read

Type 3x faster with your voice using Voibe - private, offline dictation that works in every app on your Mac

Voibe is voice-to-text for Mac that runs 100% on-device.

Hold Fn, speak, release.. your words land in any app: Cursor, Claude, ChatGPT, Slack, Gmail, anywhere your cursor is.

No cloud. No uploads. Audio is discarded the instant it's transcribed. Every other dictation tool streams your voice to a server.. Voibe runs entirely on your Mac. YOU are the server.

Sub-300ms latency. 97%+ accuracy.

Developer Mode for VS Code & Cursor. 90+ languages.

(20% off on all plans with code 8020)

When Codex Meets Blender, Magic Happens

What’s Happening AI Today

1. The US government just took Claude Fable 5 and Mythos 5 offline — globally, with no warning: On June 12, the US Commerce Department issued an export control directive that forced Anthropic to immediately block all public access to Claude Fable 5 and Mythos 5 worldwide. The takedown was instant — no transition period, no advance notice. Anthropic called it a "misunderstanding" and said it is working to restore access as quickly as possible, while issuing refunds to affected paid customers. Claude Opus 4.8 and all other Anthropic models remain fully available and unaffected. For any team that had started building on Fable 5: the capability is gone for now, and there is no public timetable for its return. This is what it looks like when AI becomes a geopolitical asset.

2. OpenAI burned $15 million per day on Sora — and shut it down after earning $2.1 million total: OpenAI has officially discontinued Sora, the AI video generation model it launched to enormous fanfare in early 2026. The reason: an estimated $15 million per day in operating costs against $2.1 million in total lifetime revenue — a ratio that made continued operation economically indefensible. The shutdown is the clearest evidence yet that the AI product market does not reward capability alone. Being technically impressive is not sufficient. Being economically sustainable is the actual test. For any team building on AI video APIs: the competitive landscape just shifted significantly, and Google's Veo and Runway are the remaining scaled options.

3. 94% of organisations increased AI spending last year — and 51% think adoption is already too fast: The Logicalis 2026 Global CIO Report surveyed more than 1,000 CIOs worldwide and found a paradox: nearly every organisation is spending more on AI, while more than half simultaneously believe they are moving too quickly. 89% describe their approach as "learning as we go." 62% have compromised on AI governance due to limited knowledge. Only 44% fully understand the risks of the AI they have already deployed. For product and tech teams, this data tells you something important: the organisations buying AI tools are not confident — they are anxious and moving fast anyway. That is both an opportunity and a warning.

Think through any hard decision in 10 minutes

Prompt: I have a decision I need to think through clearly. I am going to describe it to you, and I want you to help me think — not decide for me.

Here is the decision: [describe it in plain terms — what you are choosing between, what you know, what you do not know, and what is making it hard].

Do four things:

1. Restate the decision as you understand it — in one sentence. If I have framed it badly, say so. Sometimes the hardest part of a decision is that we are solving the wrong problem.

2. Identify the real tradeoff — not the surface one. Most decisions feel hard because of one underlying tension that we are not naming directly. What is the thing I am actually choosing between at the deepest level?

3. Tell me what I am not seeing — what assumption am I making that I have not questioned? What would a smart person who disagreed with my instinct say? What information would change my view if I had it?

4. Give me a forcing question — one question that, if I answered it honestly, would make the right choice obvious. Not a list of pros and cons. One question.

Do not tell me what to decide. Help me think well enough to decide for myself.

⚡ QUICK HITS

• Fable 5 jailbreak fully documented — the 120,000-character system prompt is now on GitHub. The technical details that triggered the government shutdown are publicly available.
• OpenAI launches self-serve Ads Manager inside ChatGPT — advertisers can now buy placements directly inside ChatGPT conversations. The ad era of AI has officially begun.
• OpenAI launches 3 real-time audio models — conversational agents, live translation, and transcription, all available via API today.
• SanDisk stock up 5,000% in 2026 — trading at $1,980, with $42B in secured contracts. AI storage demand is reshaping the memory chip market faster than anyone predicted.
• Meta buys $27B of Nebius GPU capacity — European-compliant compute for GDPR-sensitive AI training workloads, on NVIDIA's Vera Rubin platform.
• Anthropic proposes a global AI pause — Daniela Amodei floated the idea publicly. Analysts are divided on whether it is genuine or strategic positioning ahead of the IPO.

Anthropic's Valuation May Hit $1.75 Trillion

AI news highlights

• SpaceX SPCX trades today — $1.77T valuation — sets the institutional benchmark for Anthropic October and OpenAI September IPOs
• OpenAI crosses $10B annual revenue — took Salesforce 20 years; OpenAI did it in under five
• Microsoft 365 Copilot targets 1 billion seats — JPMorgan already using it across 300,000 staff, $2.5B annual value
• SK Hynix hits $1 trillion market cap — AI memory demand reshapes the chip industry
• New York passes 7 AI bills — kids chatbot safety, training data transparency, data center moratorium, surveillance pricing ban
• Microsoft ships MAI-Thinking-1 — reducing strategic OpenAI dependency, broader in-house model family
• Orion-100B: 100B parameters for $1.25/hour — AI training economics just became unrecognisable
• Claude now an iPhone option — iOS 27 multi-AI Extensions gives Claude access to 1B+ active Apple devices

Trending AI tools

🤖 Cloudskill — Govern every AI skill your team depends on. Versioned, audited, enterprise-ready.
👉 Try it →

🎨 Slashspace AI — A canvas where AI agents do the research and you do the thinking. Not a chat window.
👉 Try it →

🗣️ Kimi AI K2.6 — Browses the web, reads your files, writes code. What happens when you give it something you actually need done?
👉 Try it →

💻 VEXI — Local, multi-model coding agent with full control over which model runs which task.
👉 Try it →

📊 Langfuse — Monitor, debug, and evaluate your AI from one dashboard. Works with OpenAI, Anthropic, and Google.
👉 Try it →

🔊 Cartesia Sonic — The voice API that makes AI phone agents sound human. Most users can't tell the difference.
👉 Try it →

That’s a Wrap

SPONSOR US

Get your business in front of over 90k+ AI professionals

8020AI is the world’s #1 AI Newsletter, Read by 90k+ professionals from leading companies such as Google, OpenAI, Meta, and Microsoft.

We've assisted in promoting Over 500 AI-Related Products. Will yours be the next?

What We Can Offer:

• Launch an Advertising Campaign

• Introduce New Product or Features

• Other Business Cooperation

Get in touch today.

Or Email our founder Alamin at [email protected]

FEEDBACK

How was your experience with 8020AI today?

Login or Subscribe to participate in polls.

If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.