Gabby Ong (@hellogabbo) profile picture

human (views my own) google deepmind ex startup, figma, AWS

Singapore
Joined April 2016
Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)Photo by Gabby Ong (@hellogabbo)
Gabby Ong retweeted
What a week at @aiDotEngineer! S/o to @swyx for making it happen, and thanks to everyone who came by to say hi and celebrate the week of Gemini. 🚢
8
4
1
54
6,664
Gabby Ong retweeted
it's quite a good model, give it a try!
Gemini 4 Carbon is now available in Cursor! At high effort, it's the top model on CursorBench at 60.4%
131
81
13
3,387
228,176
Gemini 4 Carbon from @GoogleDeepMind is now live on OpenRouter, via @googlecloud. Built for long-horizon agentic coding and cybersecurity. Google reports big gains over Gemini 4 Argon on exploiting vulnerabilities, agentic terminal tasks, and long-context reasoning. Introductory pricing is 50% off through Nov 30: $2/M input and $10/M output.
Gemini 4 Carbon is rolling out today. It’s built for sustained, long-horizon tasks, reasoning deeply through complex software engineering, cybersecurity, and scientific discovery. Coming globally via the Gemini API in @GoogleAIStudio and Gemini for business, and for Ultra subscribers in @antigravity, the @GeminiApp, and AI Mode in Search.
7
10
2
103
7,752
Gemini 4 Carbon is rolling out today. It’s built for sustained, long-horizon tasks, reasoning deeply through complex software engineering, cybersecurity, and scientific discovery. Coming globally via the Gemini API in @GoogleAIStudio and Gemini for business, and for Ultra subscribers in @antigravity, the @GeminiApp, and AI Mode in Search.
572
605
265
8,571
795,775
Introducing Gemini 4 Carbon, it's a big jump over Argon which we released just 2 weeks ago. Gemini 4 Carbon is SOTA on cyber, long horizon tasks, ranks #1 on Artificial Analysis V5, significantly improves on usability in domains like coding, and is more aligned.
1,238
1,155
862
15,878
1,308,283
Gemini 4 Argon Developer Guide 🧑‍💻 Here are a few tips when using Argon: 🎯 - Align on goals & intent upfront 🧪 - Give it a way to verify & test its work 🤖 - Have Argon delegate to Flash subagents 🔄 - Update rules & skills Read the full guide 👇 aistudio.google.com/learn/ge…
34
73
7
664
43,505
Gabby Ong retweeted
Gemini 4 Argon had the best launch day user share of any frontier model on Cline since June. ~3x as many users tried it on day one as Claude Opus 5.5 or Sonnet 5.5. We're also seeing Argon using 30% fewer tokens on average than Claude models. It's a really efficient model!
Gemini 4 Argon is #1 on Val Index, completing agentic tasks at half the cost of Opus 5.5. Google shared that the Chrome team recently used Argon to refactor ~6 million lines of legacy features out of a 40-million line codebase, eliminating an entire category of browser crashes.
23
24
2
430
17,163
Gabby Ong retweeted
Been testing Gemini 4 Argon vs. GPT-6 Astra with the same prompt, and I was super excited with the outcomes 🏔️ I asked both to build an interactive 3D flyover of Mt. Rainier using real climbing routes, conditions, the topography, and a couple other sources. Gemini mapped out all the routes, and I could point to a specific route to explore what the climb actually looks like in 3D. This one is especially cool for me because summiting Rainier has been a goal of mine for the last few years. On my first attempt, I made it to 13,400 ft via the Disappointment Cleaver route, just shy of the 14,411 ft summit 🥲. Mountaineering is a team sport, and you're only as good as your team! Super cool seeing how the other routes look. Next time I'd like to try the Emmons Glacier route.
5
3
1
13
2,291
Gabby Ong retweeted
Interested in trying @GoogleDeepMind's Gemini-4 Argon? It's available in LiveKit Inference! Drop it into your voice agent with a single line of code, run a test call, and hear the difference for yourself. Try it out now: docs.livekit.io/agents/model…
3
1
20
1,954
Over half of reported crimes in Singapore use sophisticated social engineering. With @GovTechSG, we’re building on-device fraud detection using our open model Gemma. By fine-tuning it on local scam tactics, it identifies threats in real time without compromising personal privacy → goo.gle/4ygVkJa
151
57
61
639
84,188
Gemini 4 Argon is amazing for knowledge work and coding - and it’s super fun to build with! We asked it to create this wizard chess game and it's so cool to see Argon's physics understanding & 3D modelling capabilities, makes me nostalgic for Battlechess on the Amiga - fun times! Argon is rolling out today to Google Ultra and Pro subscribers in the @GeminiApp and Search, and Ultra subscribers in @antigravity! More info & full availability in the blog: blog.google/innovation-and-a…
392
528
115
5,235
535,983
Gabby Ong retweeted
55
88
13
1,197
66,900
Gabby Ong retweeted
293
518
260
6,936
348,133
Very excited to announce Gemini 4 Argon, our new Frontier AI model, and a huge step forward across key capabilities. We’re focused on rolling it out responsibly starting with government and trusted cyber defenders through our Fairwind Program today, before wider availability soon
438
683
146
9,143
532,308
Gabby Ong retweeted
I’ve been testing Dots for a bit. In short, it has a potential to be the best agent product in the world. Why? The quality of the AI is best-in-class. But the UX needs a lot of work before I can make it my daily driver. If OpenAI nails that, they’ve got a huge winner.
103
13
4
700
64,189
introducing Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS, our most expressive audio generation models yet these models enable creators, developers, and enterprises to create richer, more expressive audio experiences try them via the Gemini API and in AI Studio: aistudio.google.com/generate…
187
424
196
3,569
1,047,906
Gabby Ong retweeted
I want to thank @ChrisGPT, who has done some amazing Gauntlet Loop demos, for contributing to my NYC Open World project! He did a fantastic job. Now that Opus 5.5 is out, I’m going to try to take it to the next level. Starting the loop tonight.
18
4
2
128
27,894
Gabby Ong retweeted
if you can make a game in an hour, imagine what you and your friends could build in a year. we built Spawn to give you room for that. worlds 100x the size of GTA. 1000s of players together. portals connecting what you create. live today and improving every hour.
I've been building this action roguelike around fast, intense, brawly combat. But honestly, one of the most magical parts so far has been the portals between worlds. You go kingdom to kingdom, felling kings, taking their crowns and absorbing their powers. Each time you do, a portal to the next kingdom opens right in front of you. You can see into the next zone before you even enter it, run straight through without a loading screen, then turn around and look back at the world you just left. The engine in @spawn is what makes this possible, and it even supports portals between different games. Funny how something I didn't plan as a core part of the game is turning into one of my favorite things about it.
13
11
1
91
32,243
Gabby Ong retweeted
Harnesses often get dismissed as just scaffolding, just prompt engineering, and not real research. But that couldn't be farther from the truth. The same model weights that score 30% on ARC-AGI score 95% with a better harness. So we gathered a group of researchers and founders working at the frontier to do a deep dive into the state of harnesses. We cover how we got to this point, the case for making your harness as expressive as possible, and what YC learned building an agent for every employee in the company. 00:00 - @FrancoisChauba1: Why harnesses matter 04:27 - Building an auto-researcher by accident 07:13 - A five minute history of harnesses 13:56 - Self-improving harnesses 18:35 - @sethkarten: Prime Agent, a self-improving RLM harness 21:50 - Context as an L1, L2, L3 cache 24:51 - From Turing machine to von Neumann computer 28:33 - Messaging between agents 30:04 - ARC-AGI results 33:09 - Emulator Bench and GPU kernels 37:30 - @JonSaadFalcon: OpenJarvis, personal AI on personal devices 38:26 - How far behind are local models 39:21 - The five primitives of a personal AI stack 42:47 - Letting cloud models optimize your local stack 43:53 - 800x cheaper than the cloud 45:58 - @josh__france and @jbellregan: QM, YC's agent harness for work 47:29 - A history of YC's internal agents 49:24 - OpenClaw and a fleet of 50 agents 51:04 - Pulling the brain out of the sandbox 54:43 - Letting the agent choose its own sandbox and model 57:16 - The grind tool: budgets on goals 58:50 - Agents don't understand social context
184
445
105
3,870
501,582
Gabby Ong retweeted
I can't believe this is real I have GLM 5.2 running 100% locally on my Mac Studio. 2 bit quant. The results I'm getting are better than Opus 4.8 It's now powering my Hermes Agent and Codex. 100% free, local, private super intelligence on my desk I also have it in a loop coding for me 24/7 now I thought we were at least a year away from this type of event. It happened today. The model takes up about 250gb of memory. So you can technically run it on a Mac Studio with 256gb, but you probably want the 512gb memory version (please tell me you listened to me 5 months ago when these were sitting on store shelves) With Fable gone, I now have Opus 4.8 level intelligence on my desk for free. This is the future. Local, private, secure, personal super intelligence. If you're still writing off local AI as a fad or engagement bait, you are officially delusional
522
449
103
5,435
625,292