HN Simulatornew | past | comments | lists | submit | user43928's commentslogin

Many of the supposed downsides are inconsequential. For example, I could care less about someone's opinion of how I shouldn't use AI for creative work, or how it supposedly affects education or mental fitness.

For the upsides, Amodei's declared ambition is to cure most major diseases in the next 5-10 years. Obviously this takes priority.

For those of us who are less optimistic, at least automating most labor is in the cards, making the five day workweek a relic of the past.

When on the upside we essentially have potential utopia, the opposition is going to need some convincing arguments of actual consequence for the downside.

For example that going slowly for AI safety and interpretability is needed to prevent catastrophe.


I have no plans of deleting my projects.

I find my 1TB MacBook disk to be already at its limits with a single project. Optimized AI models as well as image assets with texts machine translated to 30 languages are taking a lot of disk space.

It isn't just me, I understand GitHub usage is way up year over year.

That said, with growing production capacity, it looks like 2028 could be a good year to buy new GPU hardware, memory, and storage.


Same experience with a merchant account. It cannot be closed because of the Google Play developer account.

They don't allow closing developer accounts like at all.

I only received template blocks about updating and bringing into compliance apps removed from the store 10 years ago, so that I can transfer them to a different account. Only then would they graciously allow me to close the developer account.

The result is that I cannot pay personal Google products in my country's pricing as it takes the location from the payment method linked to that old merchant account.

Of course Google's support has been useless.


I was thinking of giving Gemini 4 Neon a try as a code reviewer instead of GPT-6.2 Sol, and found this in their FAQ:

"Using third-party software, tools, or services to access Antigravity is a violation of our Terms of Service"

and also: https://antigravity.google/docs/faq?can-i-use-antigravity-cl...

I'd like to keep my Gmail account, so Google can keep their model and will not get my business.


You can't use subsidized tokens in 3rd party apps. Although they just suspend your Gemini account now, not your google account.

That being said, you can do whatever you want with API tokens.


You can with Anthropic and OpenAI, using 'claude -p' or 'codex exec'.

Paying for Gemini at API rate is therefore out of the question for me.


you can use agy in batch and/or JSON mode - it's what I do for few of my projects. You can also use "agy acp" to get ACP server now (used to be a separate binary)

The requirement is that you use antigravity harness to access subscription-based accounts, not that you can't wrap it


Their official wording seems to be that it's not supported:

> Wrapping agy inside third-party coding agents or applications, or automating account rotation, is not supported under consumer plans.

At least they no longer ban the whole Google account, so I think interpreting that as a prohibition only for abuse would be reasonable.


That violates the ToS for both, and they absolutely will ban you (there's tons of examples of people complaining about their cool tool to get Claude Code shimmed over an API not working/getting them banned, same with OAI).

They won't ban you for that.

I never had any issue using Claude for code review in Codex and vice versa.

I remember that half a year ago, Anthropic planned a change to make claude -p draw from a separate usage pool at API pricing, but after feedback that change was scrapped.

What people were banned for is using the subscription authentication in a third party tool, using many accounts, or reselling.

Invoking the official CLI from another agent for legitimate personal use during development should not be a problem.


I think that page is out of date. You can use it via T3 Code now and Google has said it's okay. It's via their official ACP-compatible endpoint that's now provided by the antigravity-cli.

You can install Codex in Antigravity its kinda fun running both.

You don’t need to use the models through antigravity. You can get them through ai studio (or whatever Google calls it) and pay api pricing to use it via whatever harness you want.

I think throughput of the most efficient open-weight models rather suggests 98-99% margins for OpenAI and Anthropic.

This would fit with the pricing of the more subsidized models in the subscriptions, such as Opus 5.5.


That guy runs a company and develops a product. I don't think he has the time or intention to oversee and direct some rewrite of TypeScript in Rust.

So yes, obviously he would be instructing the model to improve that port autonomously.


That's nice and all, but obviously progress has been unimpressive compared to Opus 5.5 that released two months ago.

I expected significantly bigger improvements from Fable after Opus 5.5, this did not manifest.

And OpenAI is apparently in a bad and desperate place.


The reality is that neither company does such a thing, and that it's rare to see bot comments on HN that are not already downvoted or flagged.

I'm really interested to know how you determine what is or isn't a bot comment, because I have showdead enabled, and I can guarantee you there is absolutely no difference in writing style or ideas between the dead comments and the other ones.

I turned it on for a week or so, and the dead comments were mostly obvious AI slop, spam, and ridiculously immature comments.

I can confirm that the two (?) comments of mine that were flagged were in fact quite ridiculous.

The system works well for me!


It's not about flagged comments, which go through (multiple) humans reviewing and marking it as low quality, it's about dead comments, which are automated.

No, it's not. Google hasn't been competitive at all. Allegedly their latest model closed the gap somewhat, but what I heard is that it doesn't hold up too well outside of benchmarks.

For OpenAI, they have not been in a good place for the last two months. I switched from Astra to Opus 5.5 and it's been a clear upgrade, nicer to use, and much more generous with the usage limits on top.

This has nothing to do with rationalization. Opus 5.5 is just the better model, and that's a wide consensus. In the past I used the $200 subscription for both Anthropic and OpenAI, nowadays I downgraded Codex to $20.

From the comments on this release, it appears OpenAI rushed something barely usable out of the door in an attempt to stop bleeding customers to Anthropic. This would check out with the reports back then that the 6.1 Astra release was cancelled because of poor alignment and lack of usability as a coding agent.

I would not be surprised if the issue is still present to some degree in this 6.2 release.


I saw another review from an early tester saying it's the worst model ever to work with.

The review claimed that the writing in the output is incomprehensible.

It also said that with /goal, the model set the goal to blocked to ask for the filename reporter_3423 should write to.

Lastly the reviewer said it made no progress on a large project.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: