Asked Gemini for a gag photo of me today. It asked for auth to Google Photos to pull photos of me.
Cool feature @GeminiApp
Cool feature @GeminiApp
1
0
0
49
Evan Grenda retweeted
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.
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.
1,241
1,154
15,896
1,298,897
Tru
The one MCP you should always have installed if you work on frontend
chrome-devtools-mcp lets Claude Code control a real Chrome browser
Reproduce bugs, inspect network requests, debug console errors, analyze performance, and automate browser interactions
You can ask Claude to open localhost:3000, reproduce a bug, and find what’s broken
Install it in Claude Code:
/plugin marketplace add ChromeDevTools/chrome-devtools-mcp
/plugin install chrome-devtools-mcp@chrome-devtools-plugins
The latest update lets you analyze heap snapshots without a live Chrome connection, making it easier to investigate memory leaks offline
Github:
github.com/ChromeDevTools…
1
0
0
65
Agree that workloads will run on cheaper models.
This is a reason to be bullish in Google . There will be always be a place for frontier intelligence but majority of everyday workloads (outside eng) will run on optimized models and compute. Google is the only company that demonstrate success with the full stack to converting energy > tokens + good at building products.
Plus Antigravity is good, will get better plus 3.5 pro coming 😈
This is a reason to be bullish in Google . There will be always be a place for frontier intelligence but majority of everyday workloads (outside eng) will run on optimized models and compute. Google is the only company that demonstrate success with the full stack to converting energy > tokens + good at building products.
Plus Antigravity is good, will get better plus 3.5 pro coming 😈
Good take
My guess is
- demand for intelligence is near infinite
- but 80% of workloads will be running on 99% cheaper models within 12-18 months
- 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?)
- rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though
- this leads me to think the limiting factor will be energy and compute, not better models
At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially.
0
0
2
207
Moving from slides to html presentations last year has been a game changer.
I have a local file system with a catalog of html slides (slide for knowledge worker vs C suite vs eng leader) and have built a workflow to build a narrative > select slides > tailor to them to audience in @antigravity. GUI coming next to share with the org.
Leading AI GTM at ~40 xl enterprise accounts would hard without this.
TY @GoogleDeepMind for making an awesome agent harness
I have a local file system with a catalog of html slides (slide for knowledge worker vs C suite vs eng leader) and have built a workflow to build a narrative > select slides > tailor to them to audience in @antigravity. GUI coming next to share with the org.
Leading AI GTM at ~40 xl enterprise accounts would hard without this.
TY @GoogleDeepMind for making an awesome agent harness
2
0
2
162
Peloton should pivot into Inference-aaS and ai hardware
GPUs on bike and unlimited energy supply with spin classes
Surely this makes more sense then all birds
GPUs on bike and unlimited energy supply with spin classes
Surely this makes more sense then all birds
1
0
2
117
You now need to build for agents. API and CLI tools for agents are a must.
Bonus points if it’s unambiguous so the agent uses them well.
Bonus points if it’s unambiguous so the agent uses them well.
“We’ve also been moving off legacy systems with poor, slow, outdated, and inconsistent APIs.”
If you’re building software that can’t work fully headlessly in a way that agents want to use, you’re not prepared for what the future of software is going to look like.
Agents will use software 100X more than people, and people will more and more interact with their data and workflows via agents across many different platforms.
This is the real risk but also opportunity for platforms right now. Software doesn’t go away, but it becomes the guardrails and business logic for what agents are able to operate on. But if you can’t connect to wherever the agents want to do that work, you’re DOA.
0
0
1
151
This is why cowork feels like taking a step back. Cowork pounds away on these fragile mcp servers.
What happens in Claude code: you setup mcp servers, then they break, you ask them to get fixed and over time it gets replaced with api + cli. And now everyone is coming out with CLI based tool. Especially excited about the ability to launch cloud services via CLI.
Still a pain to go in and create the api + permissions but worth it.
What happens in Claude code: you setup mcp servers, then they break, you ask them to get fixed and over time it gets replaced with api + cli. And now everyone is coming out with CLI based tool. Especially excited about the ability to launch cloud services via CLI.
Still a pain to go in and create the api + permissions but worth it.
The cofounder and CTO of Perplexity, @denisyarats just said internally at Perplexity they’re moving away from MCPs and instead using APIs and CLIs 👀
0
0
0
165
Built a voice agent to to present itself and engage with customers at NVIDIA GTC
The robots are selling for us
The robots are selling for us
2
0
4
2,581
I generally have a hard time describing what my jobs is. This is how I explain it:
1/ I sell cloud infrastructure to the companies building voice AI - the ones making the voice models, and the companies making tools to orchestrate the voice models. My job is to make sure they train their models and host their solutions on AWS.
2/ I also have a second set of customers - end customers that want to build voice agents but need guidance on how to do it. So I also have to build and sell full stack solutions and drive consumption with the end customer.
3/ But I am also helping the voice model providers and orchestration frameworks with GTM - how and what to sell because my customers are feeding more the requirements for what they are building.
How do I make it easier to understand?
1/ I sell cloud infrastructure to the companies building voice AI - the ones making the voice models, and the companies making tools to orchestrate the voice models. My job is to make sure they train their models and host their solutions on AWS.
2/ I also have a second set of customers - end customers that want to build voice agents but need guidance on how to do it. So I also have to build and sell full stack solutions and drive consumption with the end customer.
3/ But I am also helping the voice model providers and orchestration frameworks with GTM - how and what to sell because my customers are feeding more the requirements for what they are building.
How do I make it easier to understand?
1
0
0
109
@grok how do i explain this in simple terms without diluting the complexity of my job
1
0
0
92
