krat0sprakhar

> Argon will launch at an introductory price of $2 per million input tokens and $10 per million output tokens, with cached input tokens priced at 95% off input token price.

https://blog.google/innovation-and-ai/models-and-research/ge...

1/5th the price of Astra and Fable

(I work at Google) Yes, internally we all use Jetski (internal version of Antigravity). Outside of Gemini, Opus models are supported and allowed for internal use. No OpenAI models since they are not on Vertex

> The motivations of people building the AI's are not the same as the people in charge of the labs. Looking at the last few years, has been a one-way door from OpenAI to Anthropic, and the main reason does not seem to be better compensation or even them winning, but mainly the fact they advertised to these employees that they would be the most careful when building this magic lamp. Their stance around 2023 / 2024 was one of the key reasons they were able to attract this talent.

> If recent news is to be believed, Anthropic culture even today seems to lean heavily on this effect

> In my opinion it has been the leading factor in getting and retaining the best employees who are often very worried about humanity dying to super intelligence.

I had not considered this viewpoint but this makes a lot of sense. A lot of Anthropic employees truly believe this and callout emphasis on safety as a key reason they work there. Now, Dario's post "Pacing the frontier" makes even sense -- it is as much for his employees than it is for the rest of the world.

Cmon, you're being too harsh.. ELIZA was a such a fun chatbot to chat with.. I was blown away the first time I tried it in 2000s :)

TBH: I really like how fast 3.8 Flash is... Once I have clear plan, I feel quite confident in delegating large parts of implementation to Flash and Luna

Can't agree more. Between 5.6 Luna and Gemini 3.8 flash I'm so happy for the value I'm getting for my dollar (subscription pricing not API pricing) :)

Are you using the harness with openrouter? what's your preferred model provider?

Here's an example for using it with GLM:https://docs.z.ai/devpack/tool/claude ... works similarly for other models

Use antigravity CLI (https://antigravity.google/product/antigravity-cli)

The AI slop that's everywhere is making it worse. Even if someone doesn't care about AI, their feeds (YT, Instagram, LinkedIn) is now filled with AI slop

Haha, this is my favorite one so far. Well done sir

Have you tried using the free tier in https://aistudio.google.com/api-keys? Sorry for your billing issues - these can be super annoying to resolve

+1 GCP can be confusing - I totally get that.

Even if you have GCP projects, I'd still recommend the AI studio UI - easier to figure out. Also, you can easily see the free tier in AI studio and just use your API key from there.

Spend caps can now be set https://blog.google/innovation-and-ai/technology/developers-...

Can you help me understand how it is hard to get an API key from Google? You just head on over to http://aistudio.google.com/api-keys and create a key... not any different from platform.openai.com?

Disclaimer: I work in Google so it might be that this link is not publicly well known

I was at Glacier Bay National Park just last week (https://photos.prakhar.me/alaska) and it was just depressing to hear from the park rangers on how far the glaciers have receded. Really grateful to have been able to witness them in person but it is sad to see what we've done to this planet.

> Whenever I find a use for AI, it so drastically cuts the time to accomplish tasks that I’m left wanting to do more with my (finally copious) free time!

I get what the author is trying to say but this does not match my experience nor does it match the employee productivity findings from AI adoption in enterprises.

Which harness did you use kimi with? opencode?

Cambridge for sure. One aspect to consider is the funding. Would you have to get into years of debt to pay off the Masters? If no, then it is a no brainer.

Context: Been at Google for 10 years. I have a master's in CS/ML myself. Wouldn't trade it for working in FAANG a couple of years earlier.

This is awesome - I tried building something similar that my wife and I use. Do you support import from instagram? IMO that's what a lot of friends want give how much of cooking / recipe content is out there (and Instagram's search is atrocious)

I can't thank you enough for mentioning Agalloch! I used to listen to them so much in university but somehow completely forgot about them until I read this comment!

I'm going to lose myself to "The Mantle" this weekend (best part is that I can now learn to play these songs on guitar). Thank you so much again - you made my day!

Can't agree with this more. I also started learning guitar and producing music very recently. I have no interest in getting heard and appreciated (on most days atleast).

It has been a tremendously rewarding journey to create new music and see myself improve. 10/10 would do again.

Looks really interesting but is the entire project ~5K lines of Go in one file? Don't you find that hard to read or are you purely vibecoding this?

> This project is 99.9% vibe-coded on the exe.dev platform using their Shelley Web AI Coding Agent and Claude Opus 4.5. Take that as you will.

Ah, wait - should've read the README before commenting. /facepalm

Thanks for sharing the project - will try it out!

$10M in the bank is very different for a family of 4 staying in VHCOL (Bay Area/ Manhattan) vs single person staying in Bangkok. The answer to this question depends on a lot on that versus job satisfaction etc.

Do you have a tutorial that I can follow?

> then I’d wait 10 minutes, then they’d cancel. Rinse and repeat for 30 minutes,

This is such a common problem in SF (esp in odd times / from the airport). Waymo has been a lifesaver in these situations.

Looks interesting and would love to try it out on my side project, but it'd be great if you could add your own package (https://github.com/davialabs/davia/tree/main/packages/agent) as a showcase.

Checkout this thread for who those people are: https://news.ycombinator.com/item?id=44258139

> Google also licensed the Windsurf IP.

Exactly! That definitely means that the $82M ARR business and the tech behind it is definitely valuable to Google

> Not much explanation yet why GPT-5 warrants a major version bump

Exactly. Too many videos - too little real data / benchmarks on the page. Will wait for vibe check from simonw and others