gatio

Except: Enterprises are driven toward API pricing, with at most some bulk purchase discounts), because they need enterprise/security features (central control).. If an enterprise gets a 50% discount on API pricing, they are doing well (OpenAI publicly offers a 30% discount for at least $1000 in credit, so we can safely assume 30% as a starting point).

> But anthropic is the supply chain risk?

OpenAI is bending over backwards to research and win those juicy government contracts to ensure their place as a strategic supplier worthy of protection and bailouts as needed.

Seriously though, it would be interesting to learn how these ended up being targeted. Unfortunately there's lower odds of full public RCAs with governments.

Prefab & faster construction possibilities would be a big one.

Plus you can alert the crew if something's off before take off (eg. loose/missing screws/panels). A forner colleague of mine did that, and got to see the entire maintenance crew come out for an impromptu RCA and incident response.

It will typically be off by default for business plans...and enforceably off in expensive enterprise plans.

> Maybe their companies will eventually do these things, but haven’t yet.

I think they are working with customers to improve the LLMs and tools for these use-cases. They almost certainly also hire experts to help filter out nonsense, pseudo-science and help curate trusted knowledge bases for training, but it will almost certainly be the customers who deliver the major results, and the AI companies will claim some of the credit. That said, patents for important medicine might help with the bottom line, so I could imagine partnerships and JVs.

> at the very least it will be a good tool to help researchers do their jobs.

Indeed.

> Personally I’d like to see them actually start benefiting humanity by doing all the things Sam has claimed they will like curing disease, cancer, global warming, etc.

It makes more sense to leave curing disease & cancer to the experts, with tools (like AI) being developed by AI experts.

Call me crazy, but I want separate organizations and experts for medical vs finance vs space vs climate vs AI research.

+++ATH0+++ ... good times

> Perhaps there is also "firewalling" on the router and the phone can't even have incoming connections. Not sure.

Correct. Every home wifi router worth its salt will firewall incoming connections by default, whether v4 or v6. It's then possible (unless it's some shitty ISP-provided locked down device) to add specific allow rules, or allow all for a particular client.

egress is typically wide open, although sometimes they lock down particular protocols by default (eg. smtp, bittorrent)

"when given unlimited access to both" is carrying a lot there.

One may as well say "given unlimited budget".

I believe the 1M context window can be configured in Codex (config file edit). The price per token increases when >256k though, but it can be useful when compaction at-that-moment would be detrimental.

OC provided a link to the git repo for his "binary switcher", which you should investigate..

Very true. If things are too reliable, systems can come to depend on them always being so so reliable... So it can actually pay off to inject transient issues deliberately.

https://netflix.github.io/chaosmonkey/ and similar can help.

That said ... most instability is introduced with normal changes, so every engineer can be a chaos monkey. ;-)

I'd raise an eyebrow, because plumbers here are a trade with on the job training, and pipework is an essential part of it.

Software has a huge surface area these days. That a trained software engineer isn't aware that HIDs and input subsystems are often highly hackable from userspace is no surprise to me.

They may also have been leaning into a compliment of OP in a self deprecating way.

shrug

As a native english speaker, your comment came across as an unnecessary slight. It might not have been the intent.

Not everyone has had the same experiences, interests, or life/work opportunities in the same way, and this person seems genuinely interested in learning more. It's something to be encouraged.

This is very subjective. You are willing to outsource one core competency (coding), but not another (documentation and other written comms) ?

I think the bigger personal challenge is to not outsource thinking, understanding and judgement as we leverage new tools, and to continue our personal learning journeys. After 20+ years as a software engineer, you can probably coast on past experience for a long time... but if you aren't learning and exercising (one way or another), your skills will stagnate. That would be sad.

Please read over https://news.ycombinator.com/newsguidelines.html

> codex capped ~300k context when claude does 1m.

That's configurable in codex.. but there is a higher cost/usage to using it.

Did you dream of ColdStore (https://coldstore.sourceforge.net/rationale.html) ?

I would expect/hope that all egress traffic from the agents in these tests would have to go through a MITM proxy layer, which intercepts and logs all network traffic attempts.

The failure is likely in detection, where they assumed traffic to internal systems (eg. Artifactory) was inherently safe. Which is a daft assumption at this point.

I often use it to simplify & target my responses, and avoid over-sharing technical details.

At the same time, I use the Q&A cycle to improve my vetted knowledge base, which I also share (and can reference in answers).

Just the other evening there was a group of teenagers in the park across the road from my house... playing french boulles, chatting and playing Beatles songs and other great music loudly. There is hope.

Except... it does produce solid code in many scenarios. A lot of the time we are just applying the same patterns and tools to a specific problem. If you understand the problem, agents can and are producing quality code.

I do a lot with orchestrating cloud infra, provisioning and deprovisioning workflows, lifecycle management, updating CMDBs, access management, deployment automation, auditing configurations, and some APIs in front. The code is boring, mostly glue. The agents nail it quite consistently, especially with some guidance.

I do know what I want out of it, and that is deterministic quality code to solve specific business problems, so that we don't have people (or their agents) running amok.

You probably install all sorts of unverified crap on your laptop already, and since you are aware of this risk, you are in a good position to mitigate it. Sandboxes are a thing.