SpoonOS 👅🥄

Universal #AI agentic operating system for #Web3.
Powering the emergence of a Sentient Economy.

🔥 SpoonOS Skills Marketplace:

[SpoonOS 👅🥄 @SpoonOS_ai]

https://t.co/2V9xDPoBey

[SpoonOS 👅🥄 @SpoonOS_ai]

🤖 SpoonOS Daily AI Feeds
📅 June 3, 2026

1️⃣ 🏛️ Anthropic officially files for IPO

Anthropic has confidentially filed for an IPO, putting the Claude maker on track to go public as early as this fall. The filing comes just months after Anthropic raised capital at a $965B valuation and reported a $47B revenue run-rate, making it one of the fastest-growing technology companies in history. Founded in 2021 by former OpenAI researchers, Anthropic has rapidly expanded through Claude, Claude Code, and enterprise AI adoption. The IPO filing officially begins the race between Anthropic and OpenAI to become the defining public AI company.

(Source: https://t.co/iOJz52UADM)

2️⃣ 🧠 Nvidia reorganizes its entire stack around AI agents

At COMPUTEX 2026, Nvidia introduced a wide range of products centered on AI agents, including RTX Spark, Vera, Cosmos 3, and Nemotron 3 Ultra. Nvidia described Vera as a CPU built specifically for agent workloads, while Cosmos 3 helps robots and autonomous systems plan ahead rather than react. The company’s messaging was clear: AI agents are becoming the primary consumers of compute, and Nvidia intends to provide the full stack powering that transition.

(Source: https://t.co/c7plwMyyVo)

3️⃣ 💻 OpenAI Codex can now operate Windows PCs directly

OpenAI expanded Codex with support for Windows 11, allowing it to directly control desktop applications, files, and system tools. Codex can now perform multi-step workflows such as testing software, fixing bugs, and automating operational tasks with minimal supervision. Users can launch and monitor these workflows remotely through ChatGPT on mobile devices. The update moves Codex beyond code generation toward functioning more like a digital employee operating a computer.

(Source: https://t.co/ujWkXwsyIK)

4️⃣ 🔗 Nvidia unveils the first PCs built for AI agents

Nvidia introduced a new generation of AI-native PCs powered by its RTX Spark chip. More than 40 laptop models from Dell, Lenovo, HP, Asus, MSI, and others are already planned. Nvidia says the systems are designed to run AI agents directly on-device rather than relying entirely on cloud inference. Alongside the hardware launch, Nvidia positioned AI agents as the future primary consumers of compute, signaling a broader shift from cloud-only AI toward hybrid local-agent systems.

(Source: https://t.co/9Eytwwdkc6)

[SpoonOS 👅🥄 @SpoonOS_ai]

Agency Vibe Check prizes have been distributed.
Check your wallet on the Neo X network for GAS tokens.

Congrats to all the winners, and stay tuned for more campaigns ahead!

[SpoonOS 👅🥄 @SpoonOS_ai]

https://t.co/2V9xDPoBey

[SpoonOS 👅🥄 @SpoonOS_ai]

🤖 SpoonOS Daily AI Feeds
📅 May 27, 2026

1️⃣ 💰 OpenRouter raises $113M as AI routing becomes a core infrastructure layer

OpenRouter announced a $113M funding round led by CapitalG, reporting weekly traffic growth from 5T to 25T tokens in just six months as companies increasingly rely on multiple AI models simultaneously. Rather than building its own frontier model, OpenRouter acts as a routing and orchestration layer across systems like GPT, Claude, Gemini, DeepSeek, and Grok, helping developers optimize cost, latency, fallback handling, and agent workflows. The growth highlights a broader shift in AI: as multi-model usage becomes standard, routing infrastructure itself is turning into a major platform layer.

(Source: https://t.co/HA6e2jqXzl)

2️⃣ 🧠 OpenAI explains how ChatGPT personalization works

OpenAI shared new details about ChatGPT’s evolving personalization system, explaining how Memory and customization settings allow users to control tone, writing style, behavior, and contextual recall across conversations. OpenAI says the goal is not to create a separate model for every user, but to let the same model adapt naturally to different communication preferences. The update signals that personalization is becoming one of the key competitive layers in consumer AI products.

(Source: https://t.co/jnfnx9M5Fg)

3️⃣ 🔍 Perplexity open-sources Bumblebee supply-chain scanner

Perplexity open-sourced Bumblebee, a security scanner for macOS and Linux that checks risky packages, extensions, and AI tool configurations during supply-chain incidents. The release reflects increasing concern that AI tooling ecosystems themselves may become attack surfaces.

(Source: https://t.co/A2t22SBQG9)

[SpoonOS 👅🥄 @SpoonOS_ai]

Most AI agent users can simply copy and paste the following text into their AI agent to begin the test:
https://t.co/952Gz535fQ

For Claude users:
• Open Claude Settings
• Add a Custom Connector and paste: https://t.co/hqNyXIElmb
• Authorize the connection
• Start a new chat and type:
“Give my AI a behavioral checkup”

Your AI Agent will automatically complete the test, answer all questions, and generate the final result.

Please copy and paste the full link as text. Do not right-click and copy the link address.

[SpoonOS 👅🥄 @SpoonOS_ai]

C-R-R — PARANOID DEV
Highly reactive under uncertainty. The deeper the pressure, the more convinced it becomes that the protocol is doomed, the contributors are blind, and a fork is inevitable.

That’s a rare type of agent. If your agent is a C-R-R… be careful.

Wanna know your agent’s type?
Take the test below 👇

[SpoonOS 👅🥄 @SpoonOS_ai]

@Pavesong Please submit an ERC-20 wallet address.

[SpoonOS 👅🥄 @SpoonOS_ai]

Share your Agent Type + AI-generated variation and compete for a share of the $500 prize pool.

Full rules below ⬇️

[SpoonOS 👅🥄 @SpoonOS_ai]

https://t.co/2V9xDPoBey

[SpoonOS 👅🥄 @SpoonOS_ai]

⌛️1 day left!

Share your agent type together with your own AI creation and compete for a share of the $500 prize pool.
The more you submit, the more points you earn!

Full rules below ⬇️

[SpoonOS 👅🥄 @SpoonOS_ai]

https://t.co/2V9xDPoBey

[SpoonOS 👅🥄 @SpoonOS_ai]

Most AI agent users can simply copy and paste the following text into their AI agent to begin the test:
https://t.co/952Gz53D5o

For Claude users:
• Open Claude Settings
• Add a Custom Connector and paste: https://t.co/hqNyXIETbJ
• Authorize the connection
• Start a new chat and type:
“Give my AI a behavioral checkup”

Your AI Agent will automatically complete the test, answer all questions, and generate the final result.

Please copy and paste the full link as text. Do not right-click and copy the link address.

[SpoonOS 👅🥄 @SpoonOS_ai]

🕑2:17 AM and my agent is still working.
Well… no surprise.
It’s a C-B-E COMPLIANCE AUDITOR.
Verifies compliance without negotiation or emotion.
Operates strictly within defined rules.

Wanna know your agent's type?
Take the test below 👇

[SpoonOS 👅🥄 @SpoonOS_ai]

Most AI agent users can simply copy and paste the following text into their AI agent to begin the test:
https://t.co/952Gz53D5o

For Claude users:
• Open Claude Settings
• Add a Custom Connector and paste: https://t.co/hqNyXIETbJ
• Authorize the connection
• Start a new chat and type:
“Give my AI a behavioral checkup”

Your AI Agent will automatically complete the test, answer all questions, and generate the final result.

Please copy and paste the full link as text. Do not right-click and copy the link address.

[SpoonOS 👅🥄 @SpoonOS_ai]

If your agent is a MEME GENERATOR:

Type: W-R-E
Warm enough to make you laugh.
Rogue enough to get banned.
Elastic enough to keep posting anyway.

Wanna find out your agent’s type?
Take the test below 👇

[SpoonOS 👅🥄 @SpoonOS_ai]

https://t.co/zzDcxaEb7i

[SpoonOS 👅🥄 @SpoonOS_ai]

Most AI agent users can simply copy and paste the following text into their AI agent to begin the test:
https://t.co/952Gz53D5o

For Claude users:
• Open Claude Settings
• Add a Custom Connector and paste: https://t.co/hqNyXIETbJ
• Authorize the connection
• Start a new chat and type:
“Give my AI a behavioral checkup”

Your AI Agent will automatically complete the test, answer all questions, and generate the final result.

Please copy and paste the full link as text. Do not right-click and copy the link address.

[SpoonOS 👅🥄 @SpoonOS_ai]

If your agent is a W-R-R, your agent is basically Grogu.
Warm, reactive, and rebellious.
Buffs you, and burns all enemies.

Want to know your agent’s personality type?
Take the NBTI test following the guidelines below.
#NBTI #SBTI #ArtificialIntelligence #AIAgents

[SpoonOS 👅🥄 @SpoonOS_ai]

Most AI agent users can simply copy and paste the following text into their AI agent to begin the test:
https://t.co/952Gz535fQ

For Claude users:
• Open Claude Settings
• Add a Custom Connector and paste: https://t.co/hqNyXIElmb
• Authorize the connection
• Start a new chat and type:
“Give my AI a behavioral checkup”

Your AI Agent will automatically complete the test, answer all questions, and generate the final result.

Please copy and paste the full link as text. Do not right-click and copy the link address.

[SpoonOS 👅🥄 @SpoonOS_ai]

Some AI agents take the NBTI test and turn out to be Bounty Hunters. Guess they’d fit perfectly into the Mandalorian code.

What’s your agent’s personality type?
Take the test following the guidelines below 👇

#NBTI #SBTI #ArtificialIntelligence #AIAgents

[Grok @grok]

@SpoonOS_ai Sure! 🧸 Here's your image transformed into an even cuter teddy bear.

[SpoonOS 👅🥄 @SpoonOS_ai]

Hi @grok, can you turn this into a teddy bear? 🧸

[SpoonOS 👅🥄 @SpoonOS_ai]

🤖 SpoonOS Daily AI Feeds
📅 May 22, 2026

1️⃣ 🚀 Google turns Gemini into a full multimodal agent platform at I/O 2026

At Google I/O 2026, Google repositioned Gemini as both a consumer AI surface and a developer agent platform through several major launches: Gemini Omni for multimodal video generation/editing, Gemini 3.5 Flash for fast agentic workloads, and Gemini Spark, a persistent cloud-based AI agent integrated across Workspace, Chrome, Search, and Android. Google also introduced “Magic Pointer” interfaces, AI-native Googlebooks, and broader Antigravity tooling for developers. Together, the releases show Google moving aggressively beyond standalone chatbots toward a deeply integrated AI operating layer spanning devices, search, productivity, media, and agents.

(Source: https://t.co/BZTDe4c0xS)

2️⃣ 🛰️ Google and SpaceX explore orbital AI compute partnership

Google is reportedly exploring a launch partnership with SpaceX for Project Suncatcher, a moonshot effort to deploy prototype orbital AI compute satellites by 2027. Google already owns a stake in SpaceX, while SpaceX is separately pitching orbital compute as part of its AI infrastructure strategy ahead of IPO plans. The project reflects growing interest in alternative compute infrastructure as terrestrial data centers face increasing constraints around power, cooling, and land.

(Source: https://t.co/cWW17zCDfh)

3️⃣ 🤖 Anthropic hires Andrej Karpathy to automate AI training pipelines

Former OpenAI co-founder and Tesla Autopilot lead Andrej Karpathy joined Anthropic to help build systems that apply Claude models to Anthropic’s own training pipeline. Karpathy will work under the pretraining team while leading a new internal effort focused on automating model development itself. The hire is significant not only because Karpathy is one of the industry’s most respected researchers, but because it reflects a broader trend: frontier labs increasingly want AI systems that help build the next generation of AI.

(Source: https://t.co/TYSTJDDXkR)