There will be no “jobs apocalypse” due to AI — but there will be job chaos. Our 2025 AI Job Impacts Analysis found that starting in 2028-2029, AI will create more jobs than it eliminates. Yet, each year, over 32 million jobs will be significantly transformed. Explore and plan for the four scenarios for human workers in the age of AI: https://t.co/bt8JM0jCNI #AI #Jobs #ArtificialIntelligence

Never a better time to get Gemma4 running in OpenClaw https://t.co/UP4MQmAB6O

Never a better time to get Gemma4 running in OpenClaw https://t.co/UP4MQmAB6O

Google has released Gemma 4, four open weights models with multimodality support. The flagship 31B model (39 on the Intelligence Index) uses ~2.5x fewer output tokens than Qwen3.5 27B (Reasoning, 42) but trails it by 3 points on intelligence @GoogleDeepMind's Gemma 4 includes four sizes: Gemma 4 31B (dense, 39 on the Intelligence Index), Gemma 4 26B A4B (MoE, 4B active, 31), Gemma 4 E4B (8B, 19), and Gemma 4 E2B (5.1B total, 2.3B active, 15). Gemma 3 was instruct-only at 27B, 12B, 4B, 1B, and 270M; Gemma 4 adds reasoning mode, native video and image support across all sizes (with audio input for Gemma 4 E2B and E4B), doubled context windows, and Apache 2.0 licensing. The nearest open weights models by intelligence to the 31B are Qwen3.5 27B (Reasoning, 42), GLM-4.7 (Reasoning, 42), MiniMax-M2.5 (42), and DeepSeek V3.2 (Reasoning, 42). Qwen3.5 also supports images and video natively; DeepSeek V3.2 and MiniMax-M2.5 are text-only. Key benchmarking results for the reasoning variants: ➤ Gemma 4 represents a large intelligence jump over Gemma 3. Gemma 4 31B (Reasoning, 39) is +29 points over Gemma 3 27B Instruct (10), Gemma 4 E4B (19) is +13 points over Gemma 3n E4B Instruct (6), and Gemma 4 E2B (15) is +10 points over Gemma 3n E2B Instruct (5). Context windows also doubled from 128K to 256K for the larger models, and increased 4x from 32K to 128K for E2B and E4B ➤ Gemma 4 31B (Reasoning, 39) trails Qwen3.5 27B (Reasoning, 42) by 3 points, primarily due to weaker agentic performance. On non-agentic evaluations, the models are more competitive: Gemma 4 31B leads on SciCode (43% vs 40%) and TerminalBench Hard (36% vs 33%), while scoring similarly on GPQA Diamond (86% vs 86%), IFBench (76% vs 76%), and HLE (23% vs 22%) ➤ Gemma 4 31B is notably token efficient, using 39M output tokens to run the Intelligence Index vs 98M for Qwen3.5 27B (Reasoning). This is ~2.5x fewer output tokens for a model scoring 3 points lower. For context, the other models at the 42-point intelligence level also use significantly more tokens: MiniMax-M2.5 (56M), DeepSeek V3.2 (Reasoning, 61M), and GLM-4.7 (Reasoning, 167M) ➤ Gemma 4 26B A4B (Reasoning, 31) activates just 4B of its 27B total parameters and is ahead of select peers in the ~3-4B active parameter range. Qwen3.5 35B A3B (Reasoning, 37) leads models with ~3B active parameters and is 6 points ahead of Gemma 4 26B A4B, with notably stronger agentic capabilities (Agentic Index 44 vs 32). GLM-4.7-Flash (Reasoning, 30) scores slightly lower than Gemma 4 26B A4B with 3B active parameters ➤ The smaller Gemma 4 E4B and E2B models perform better on AA-Omniscience than the larger Gemma 4 variants. Gemma 4 E4B scores -20 on AA-Omniscience and Gemma 4 E2B scores -24, both substantially better than Gemma 4 31B (-45) and comparable to or better than much larger models like DeepSeek V3.2 (Reasoning, -21). The larger Gemma 4 models' AA-Omniscience scores are in line with Qwen3.5 27B (-42) and Gemma 4 26B A4B (-48) ➤ Gemma 4 E2B has 2.3B active parameters and 5.1B total, designed for on-device deployment. In 4-bit quantization, the model weights fit in under 3GB of RAM, making it suitable for background tasks, basic function calling, and multimodal understanding on mobile and edge hardware Key model details: ➤ Context window: 256K tokens (31B, 26B A4B), 128K tokens (E4B, E2B). ➤ Multimodality: All models support text, images, and video input. E2B and E4B also support native audio input ➤ License: Apache 2.0. Gemma 3 models are available under a "Gemma Terms of Use" license ➤ Size/Parameters: 31B dense, 27B total/4B active (26B A4B MoE), 8B (E4B), 5.1B total/2.3B active (E2B) ➤ API availability: The two larger models are available for free on Google AI Studio. There are several third-party providers hosting the larger Gemma 4 variants such as @novita_labs, @LightningAI, and @parasailnetwork

Just 1,500 feet (457 m) from the Artemis II launch, we deployed our special high-resolution, slow-motion camera to record the blast-off in jaw-dropping detail at 2,000 frames per second. Find out how we did it: https://t.co/MxKSqR5zes https://t.co/3L1OMVC50H

https://t.co/2lHJQGAFT0

The Investor Report for NVIDIA. (I'm meeting in the morning with their venture arm, Inception). My AI reads 600,000 posts in minutes (it is a new kind of AI from Levangie Labs). And analyzes them. And tells me what's important so I go into an important meeting very prepared. I'm sharing it with you so you have the same super powers I do. https://t.co/fOPb3makhq

Time to move to open or local models from Hugging Face! All instructions are here: https://t.co/0w0pQb87Le

Mac Mini + eGPU. Both NVIDIA and AMD supported. https://t.co/CIcIF3j7Ol

Oh no! Just got the email... @AnthropicAI kicks @openclaw out, even paid users Good thing we wrote the guide to keep your claw alive with open models! https://t.co/blUUhwMPEG https://t.co/zA6NHQ1Hyy

@zarazhangrui I distilled everyone: https://t.co/kiuZ7QXLzb All 50,000 in AI community here on X

My AI says Hermes is the winner: https://t.co/kiuZ7QXLzb after Anthropic shut down the free Claws. Or is it just a sycophantic parrot of what everyone on X is saying? :-)

この狭さで抜刀できるのヤバすぎる😳 https://t.co/yhvyxe5A9F

Awesome hosting @ivanleomk developer experience at @GoogleDeepMind at @convex for lunch. Excited about future collabs for builders. https://t.co/ZLIM2dDPqs

Time to move to open or local models from Hugging Face! All instructions are here: https://t.co/0w0pQb87Le

Time to move to open or local models from Hugging Face! All instructions are here: https://t.co/0w0pQb87Le

BREAKING: New bartowski Gemma-4 26B-A4B-it MoE GGUF Just Dropped 🤯 Dropped the IQ4_NL GGUF of Google’s Gemma-4 26B-A4B-it MoE (~26B total / ~4B active)! Bartowski’s GGUF Setup: 👇🏻 🧠Revised & quantized with llama.cpp imatrix 🚀Quant: gemma-4-26B-A4B-it-IQ4_NL.gguf 14.70 GB 💻Full 256K context window Native (text + vision) MoE Performance Wins: 👇🏻 🏆 Efficient Mixture-of-Experts (8 active / 128 total) 🚨 Clean, accurate tool calls with no overthinking 🤯 Noticeably stronger agentic workflow MoEs (beats Qwen 3.5-35B-A3B in tool-use precision) 🤖 Built for local Hermes Agent / agentic meta Quants from IQ4-XS to Q4_K_M:👇🏻 🔥IQ4_NL (14.70 GB) — best accuracy/speed balance 🔥IQ4_XS (~14.2 GB) — lightest high-quality option 🔥Q4_K_S (~15.8 GB) Q4_K_M (~17 GB) Try It 👇🏻 https://t.co/tRBEetziSo

This guy is BEYOND CRACKED. Gemma 4 already on MLX, bro has uploaded all models with quantization. 125 models uploaded in last few hours 🤯 New mlx-vlm repo also supports turbo-quant, and rf-detr too (among other things) If you are a mac dev, you better be jumping at this. Bookmark him, turn his notifications on, sponsor his work.

AI gave people the ability to build anything. But not something beautiful. Introducing Clovr. Create beautiful, ready-to-use frontends and UI from a prompt. https://t.co/a3RrFBmmM4

I built a personal OS in @GoogleAIStudio with just a few prompts - build your own apps with their own icons and see them on an mac/ios view. Everything is backed up by a firebase database too with our full stack deployments :) https://t.co/P7VhevsCgV

🚀 Hermes Workspace now connects to ANY local model Just shipped: plug in Ollama, LM Studio, vLLM, or any OpenAI-compatible endpoint — full workspace with sessions, memory, skills, file editor, and dashboard. No cloud API key needed. Your model, your machine, full agent workspace. Setup: provider: ollama model: qwen2.5:0.5b custom_providers: - name: ollama base_url: http://127.0.0.1:11434/v1 Two modes: → Portable: point workspace directly at any /v1/completions endpoint → Enhanced: route through Hermes gateway for sessions + memory + tools 6 bugs fixed in one sprint: • Startup screen no longer blocks non-Hermes backends • File editor actually loads files now • Dashboard shows real session data • Jobs API crash fixed • Dark mode text visible • Local model provider routing works end-to-end GitHub: https://t.co/zf3xrKYFqZ Site: https://t.co/z3BsDUGNcv

+1 more

cursor now has design mode (⇧+⌘+D) - click to edit, drag to draw - shift + drag to box things in - add directly to chat with ⌥+click https://t.co/nh7RXNV9LU

It’s over. Officially. No more Claude in OpenClaw. Way to drop this Friday late afternoon @AnthropicAI So lame https://t.co/bzICltgr3R

Codex is great and a good alternative to Claude Code but Apple doesn't position Apple as an alternative to Samsung or cheaper Samsung or better Samsung or anything like that. OpenAI needs to get its comms together without employees diluting the brand in random ways. And its very straightforward to see Anthropic's comms with only a few folks posting about anything related to Claude Code.

This was a pretty fun idea to hack on, got it working on a simple prototype. Time to hook up some extensions and I'm good to go https://t.co/8ynjyVsJyK

The new Grok Imagine model is particularly good when you use the LLM to help write or refine prompts ✨ I often start with something simple, take the output to Chat, and ask for help iterating on an edit or enhancement. You end up with something much more detailed! https://t.co/184FKsgFFV

This is how nations commit suicide. https://t.co/xTD4EwEkzr

🇺🇸 @nickshirleyy drops Elon’s core lesson: “You’ve gotta do what you think is right.”' He calls it "truthmaxxing”: trusting your own judgment above everything. Pure Elon: follow your conviction, no matter what. https://t.co/Q9KcWhrolR

Pro tip. Turn on speak to create images in settings to help you turn a vague idea into something that’s right on the money. https://t.co/rubeBi1ZFt

ANTHROPIC HAS OFFICIALLY BANNED OPENCLAW USAGE!!! no this is not ai-generated Starting April 4, users will no longer be able to use their Claude Subscriptions for third-party harnesses including OpenClaw. THIS IS CRAZY. 👇 https://t.co/TrVI64LucJ

@VadimStrizheus @Scobleizer False https://t.co/R043Gsp5RI