ubermon

For some frontier models like Fable 5 it doesn't matter, but for models less trained on long horizon tasks it very useful.

I think it is more on distrusting python/typescript in general. Those dynamic scripting language are too easy to poison.

we need a real agent that represent our own interest and deal with a more powerful general agents. And we can delegate answering this question to those more personal agents.

lol same here, this is what I use https://github.com/AntigmaLabs/ante-preview

I think this is mostly being overengineered. At this stage, a curated MEMORY.md is enough: short, readable at session start, and easy to correct when it becomes stale.

The session history is another database like any other database, can be used or invoked by agent like one.

Both scores come from a fixed fallback rule: Opus 4.8 and Sol xhigh run first; games scoring below 80 are rerun with Fable 5 and Sol max, respectively, and the higher per-game score is retained.

hmm, this is like pass@n until you get the high watermark? How would this mean anything?

Thanks!

Totally agree! I think we are very early on discovering the full potential of local models

Thank you! I think there is a lot to dive deep later with different hardware, inference engine, prompt/harness setup etc.

Not yet, will conduct a more comprehensive one later

and Why We Can't Have Nice Things): A Story

very good articulation on the problem. My bet is that it is probably fine in the sense that the mastery itself is not going to be relevant.

I want to view AI coding as invention of new coding tools, at least in the way you described. I (hope) it will be more like punch-card->assembly-> BASIC/C -> scripting language -> some sort of well-structured natural language.

Give Copilot a try, it completely shift coding experience in Rust. Especially this:

I always have to search libraries and how to do things.

Once you pass the initial curve with crutch like Copilot, then you can be almost as productive (if not more, considering refactoring and testing) with your native first coding language

haha was thinking the same. The other day openAI's api hiccup gave me a small panic attack

Moat does not come from compile time but runtime. The company, the operation, and the accumulated data, the brand, the trust and the reputation.

you seem really uninformed about what is happening in this space. Just to name a few, incredible fast settlement with lending and yielding; Automatic Market Maker; Decentralized Exchange guaranteed with no wash trade, insolvency and max transparency, etc. Those defi projects are pretty much all killer apps. Hard to believe you have been in financial industry to not recognize those solid use cases and how they eliminate the middle man and premium paid for trust.

Curious on what is the general view of Indian population, twitter is full of trolls and paid campaign. From what I gather in chinese forum, I don't think any one believes China was the aggressive side, their most aggression is on getting Taiwan/ROC back.

It is funny to see on twitter some people accusing other side's aggression yet boasting about victory of getting further across LAC.

I think two important factors in play here, one is like what others said about better trust on the legal system in US to be relatively more independent; second I guess is just US market is too lucrative and significant to just let go.

I remember using this technique to get US content on Netflix while sitting in Canada

I think not manager but peers. They are mostly over achievers and somewhat competitive.

haha very possible, but it is still quite shocking to me that my mom, who is almost 60, keep telling me how fun it is...

Fueled by a recent Indian-China border conflicts, a political decision.

Could be some over zealous employees inspired by recent Indian-China conflicts and TikTok ban and decided within their own org to do it without realizing how big of a news it would be.

"The same cannot be said about other Chinese companies and similarly positioned companies in other countries. "

What are other companies? Those that doesn't pose a threat? like those ones only produce cheap toys and clothes?. I think as long as Chinese government remain as independent and "different", anything comes out of there that really challenges current status quo would receive similar criticism like yours, regardless what or how those company behaves. Curious how do you recognize those companies "need to integrate themselves with the CCP/PLA/etc"?

Surprised nobody talked about Fb's Lasso which intends to "copy" TikTok and recently got killed.

https://techcrunch.com/2020/07/01/lasso-facebook-tiktok-shut...

They are/were planning to IPO, and their financial will be published so I doubt the conspiracy theory. I've used the original TikTok(Dou Yin), it is super addictive, even my parents fall into that. They do a very good job in terms of engaging both the viewer and content producer(profit cutting etc.)

In my opinion, are still "evil" in terms of hijacking our brain, but I am a bit fed up with those prevailing political prejudice nowadays for anything related to China.

or simply doesn't know he could sell higher. Like I just have the (often wrong) mentality that electronic gadgets just deprecate right after you bought them. Quite surprised when learned that some of them worth more than retail price at secondhand market.

Is coincidence that I found most outrageous comments are by throw away accounts? I expected much more for hacker news

Do you really care how many died or do you only want to have a villain to hate? Or just your insecurity is stirred?

I am quite surprised to see those stereotypes on hacker news . Looking at all the accounts with names like systemxxx and throwawayxxx, I feel conspiracy theory is no joke