Right, how can anything that mentions grading, not even mentioning actively cheating, is allowed out of the door.

I care not so much about wrong answer, that's just another try and it gets it right eventually, but this cheating.

The fact that models detect being benchmarked and change behaviour is a failure in itself. Reminds me VW scandal with emission rules.

But what really caugh my attention is DS4 calling the openrouter with it's key. That's pretty bad for both the model and the evaluation sandbox.

Alignment? What alignment lol

DS4 when asked about censored stuff through api believed it's Claude. Otherwise it doesn't know what model it is (or pretends to not know).

That's a known playbook. Stir up shit, then create an uplausible denial which will take experts some time to debunk. Once it's figured out the story was bs, put out another three. By the week end there is 10 narratives and nobody except the field experts is sure that it's the usual suspects doing the usual things, but nobody listens to those

Examples of this happening in the past?

I do this in my own harness -- when the context grows too big it triggers to seamlessly handover to the next session of the same agent and organize the memory. A sample of most-relevant stuff becomes the next system prompt.

At some point I stopped noticing it doing it.

You need an app to login to the website. Then they say nobody uses the website

Thai banks at least have the fully staffed branches in every mall . European banks don't anymore

How do you feel about government regulation of slavery? Surely some people would benefit from having stable for-life emplpoyment instead of current unpredictable job market.

>If Kalshi is gambling, how does it differ from trading traditional financial instruments that are affected by unknown future events like Federal Reserve interest rate decisions,

Easy. Financial instruments are used to tune the resource allocation in capitalism. Gambling however is used to have fun with an idea of getting free money. As such, one is at odds with the religious beliefs based on not having fun for no good reason.

Well, if we are bringing religion into it, the Christians' problem with gambling mostly boils down to the medieval idea of work (an idea that itself was derived from Judaic philosophy at the time). In short, Work (an inherently good thing that even God engaged in) is about creating something of value. Money was the natural reward for what was made. Thus, commerce (at the time) was about circulating manufactured goods.

The ancient issue with gambling is that neither the gambler nor any operator is producing anything of value. With nothing to show from any party, medieval scholars would declare gambling undeserving of any financial flow. Most Christians are still (at least unconsciously) working from these principles, even if the operations of the economy has changed since then.

>Most Christians are still (at least unconsciously) working from these principles, even if the operations of the economy has changed since then.

Vibes and experiences are bigger part of the economy, this is true, but gambling is rightfully in the same cathegory as alcohol, (other) drugs, porn, algorithmic social media feeds etc. Some constrains are needed for all of the above, as they all create ill effects either on a person engaging in it and often on a society at large.

I think that constrains on what kind of fun you can have is big part of what religion is useful for.

> it is in the interest of banks to share costs and provide a single unified technical API, since they all have to spend the money anyway.

it would be if wasn't for the obvious reasons of those being banks