Discover • Tech News

❎ Musk Announces "Macrohard," a Joint Tesla-xAI AI Agent

Elon Musk announced on March 11 that Tesla and xAI have formally joined forces on an AI project called "Macrohard" or "Digital Optimus," a system designed to control computers and perform office tasks in real time autonomously.

The announcement, made via a post on X, marks the first concrete product to emerge from Tesla's controversial $2 billion investment in xAI—and arrives just days after reports that the Macrohard project had stalled inside xAI due to leadership departures and internal turmoil.

♊️ Gemini Is Now Built Right Into Google Workspace

Google has embedded Gemini across Google Workspace apps. The AI can now crunch your spreadsheets, write and polish documents, design slides, and dig through files on Google Drive.

The assistant lives natively in each app—it fully understands the context of whatever you're working on, with no extra uploads or tab-switching required.

Currently rolling out in the U.S. only, in English, for paid subscribers.

Would you trust an AI with your documents?

❤️ — Absolutely!
👎 — No, it'll just make a mess...

🪰 Eon Systems Shows Off First Whole-Brain Simulation in a Virtual Body

Eon Systems, a San Francisco startup, has pulled off what it calls the world's first whole-brain emulation of a fruit fly—one that controls a physically simulated body and produces a range of natural behaviors, all without any machine learning training.

The company announced on March 6, and researchers in the field are already describing it as a major step forward in brain emulation: a breakthrough that finally closes the gap between mapping a brain and actually getting it to do something.

👐 Maybe people don't actually want to book trips or buy things inside chatbots

OpenAI has stepped back from the idea of building travel bookings and shopping directly into ChatGPT. Instead, users will be redirected to third-party services and partner mini-apps.

Analysts think OpenAI doesn't want the headache: payments, real-time price tracking, cancellations, refunds, and customer complaints.

At scale, that's a lot of moving parts. And the commission margins on travel bookings are not significant enough.

🎮 Roblox Brings AI Chat Censorship

Now, instead of "####" replacing bad language, an AI will rephrase offensive messages into a more respectful form.

The message will also be marked as AI-edited.

Have you ever played Roblox?

❤️ — Yes, of course!
🤔 — Nope...

👐 OpenAI Releases GPT-5.4

OpenAI has introdused GPT-5.4 just one day after the previous GPT-5.3.

The model features a context window of up to 1M tokens and a high level of agentic capabilities: GPT-5.4 is natively trained to control a computer and outperforms Opus 4.6 and Gemini 3.1 Pro in benchmarks.

According to the developers, when creating presentations and working with spreadsheets, the model approaches the level of a junior banking employee.

➡️ GPT-5.4 is already available in our bot.

How to Try?

🔴 Go to @GPT4Telegrambot
1️⃣ Activate the PREMIUM package in the /premium section.
2️⃣ Select GPT-5.4 in the /model menu.

🖥 South Korea Warns Iran War Could Disrupt Global Chip Supply

South Korea's semiconductor industry is sounding the alarm over possible disruptions to key chip manufacturing materials as the U.S.-Israel conflict with Iran enters its sixth day.

Ruling party lawmaker Kim Young-bae, speaking after meetings with Samsung Electronics executives and trade groups, said officials warned that "semiconductor production could be disrupted if some of these key materials cannot be sourced from the Middle East."

South Korea produces around two-thirds of the world's memory chips, and the ongoing conflict risks pushing up energy costs and material prices across the industry.

💬 OpenAI Launches GPT-5.3 Instant

The company says the model is "less cringe" than GPT-5.2 and has fixed the chatbot's tendency to sound preachy, overly cautious, and, at times, overbearing. The new model is available in ChatGPT starting March 3.

🐉 Game of Thrones Movie in Development

Warner Bros. is preparing a "Game of Thrones" movie with "House of Cards" showrunner and "Andor" writer Beau Willimon writing the screenplay.

The story could revolve around Aegon I, the legendary Targaryen king who spawned a dynasty.

Of course, whether the "Game of Thrones" movie sees the light of day depends on many factors. Not everything "Thrones" that enters development gets greenlit. HBO developed several spinoffs with Martin before deciding to make "House of the Dragon" and eventually "Knight of the Seven Kingdoms."

💣 Amazon Confirms Three Data Centers Damaged by Drone Strikes in the Middle East

Amazon confirmed that three of its cloud data centers in the Middle East were damaged by drone strikes, marking the first time a major U.S. technology company's infrastructure has been knocked offline by military action. Two facilities in the United Arab Emirates sustained direct hits, while a third in Bahrain was damaged by a strike "in close proximity," according to an update posted to the AWS cloud service dashboard.

📱 Apple Introduces New Budget iPhone 17e

Apple has launched the iPhone 17e, the latest affordable model in the iPhone 17 series.

This new phone now starts with at least 256 GB of storage (twice as much as the previous version at the same price) and has a bright 6.1-inch OLED Super Retina XDR display. It brings back the simple classic notch at the top of the screen instead of the Dynamic Island. It has just one main camera—upgraded to a powerful 48MP Fusion lens with 2x zoom.

The iPhone 17e comes in three colors: black, silver, and pink. It starts at $599. Pre-orders open on March 4.

🔥 Anthropic Refused to Give the Military Unlimited Access to Their AI

Anthropic CEO Dario Amodei expressed concern that the government could use AI for "mass surveillance" and "fully autonomous weapons," and said such demands are impossible. According to him, today's systems are not ready for that.

As a result, Donald Trump ordered all federal agencies to stop using Anthropic AI. The company was labeled a "supply chain risk," which automatically bans any other Pentagon contractors from using Anthropic products in their work.

✅ OpenAI took Anthropic's place. Even though OpenAI CEO Sam Altman says the deal keeps the ban on surveillance and autonomous weapons, there is a big difference. OpenAI agreed to use AI "for all lawful purposes," and any disagreements will be settled under "specific legal and political rules." Anthropic's position meant the company itself would decide how to understand the limits.

💡 So what does this mean?

The "AI revolution" is changing not just IT, finance, and education—the military is also quickly learning to use these new tools. And the future of AI startups will probably depend even more on how willing they are to work with the government.

AI in the hands of the government...

❤️ — will bring benefits
🔥 — is dangerous for citizens!

💬 OpenAI Raises Record $110B from Amazon, Nvidia, SoftBank

OpenAI announced on Friday a $110 billion funding round at a $730 billion pre-money valuation, the largest private financing deal ever recorded. The round includes $50 billion from Amazon, $30 billion from Nvidia, and $30 billion from SoftBank, with additional financial investors expected to join as the round progresses.

♊️ Nano Banana 2: New "AI Photoshop" Is Now on @GPT4Telegrambot

Google just released its new AI image editor Nano Banana 2, also known as Gemini 3.1 Flash Image Preview.

The model keeps details and faces much better than the basic version when editing, and its detail level comes close to Nano Banana Pro.

You can try the model for free in our bot:

🔴 Go to @GPT4Telegrambot → Image Generation → Nano Banana 2

⚡️ 25 generations per week for free. Up to 200 a day with Premium subscription.

How do you like the model?

❤️ — Looks great!
🤔 — Not much progress...

📸 Instagram Will Alert Parents if Teens Search for Suicide

Instagram is adding a new warning feature for parents. The app will notify them if their teen repeatedly searches for words related to suicide or self-harm in a short time. The alert will also provide helpful guidance for parents on how to talk to their kids about these issues.

This feature starts next week for accounts with parental controls in the U.S., UK, Australia, and Canada. More countries will get the update later.