Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Trae: An AI-powered IDE by ByteDance

trae.ai

Post reply on HN
Print view View original

361–370 of 428 posts

Re: Trae: An AI-powered IDE by ByteDance

#361

Post by TeMPOraL »

codr7 wrote: post #340
falcor84 wrote: post #274

Earlier quoted context omitted.

I don't know how you compare, but ChatGPT's English grammar and vocabulary are significantly better than mine. And when I prompt it appropriately, it also seems to be a better creative writer than I am, at least for short pieces.

Except it has no freaking idea what it's doing, that's the difference.

It's giving a reply that "sounds right to a human" in context. Which, on small scale, is exactly what they, me and you are also doing when writing (or speaking), except for the infrequent cases when we force ourselves to reason through stuff very. slowly.

(This is why I believe LLM performance is best judged against human inner voice/system 1 reasoning, not the entirety of human thinking. When thinking with system 1, people don't really have an idea what they're doing either - they're just doing stuff that feels right.)

Also note that "sounds right to a human" is literally the loss function on which LLMs are trained, so between heaps of training inputs and subsequent extensive RLHF, the process is by its very construction aiming optimizing for above-human-average performance across the board.

Re: Trae: An AI-powered IDE by ByteDance

#362

Post by latexr »

CryptoBanker wrote: post #352
latexr wrote: post #343

Earlier quoted context omitted.

We’re talking about going from fully innocent and legal action to full on death penalty in just over two terms (depending on country). That seems quite a stretch for many countries where the death penalty was abolished. https://en.wikipedia.org/wiki/Capital_punishment_by_country It doesn’t seem plausible that in countries where even euthanasia isn’t allowed things could get so out of hand that the death penalty would…

You are literally debating something that hasn't happened yet. The death penalty example was just that, an example

Clearly the death penalty was not just an example, seeing as the reply from that same user made a good point about their own country becoming a war zone.

https://news.ycombinator.com/item?id=42803165

I recommend you stop assuming what other people meant and nitpicking for no discernible reason. The only one debating here is you, and you’re doing it with yourself.

Re: Trae: An AI-powered IDE by ByteDance

#363

Post by redcobra762 »

spencerflem wrote: post #327
descendant wrote: post #325

Earlier quoted context omitted.

I would like to highlight the small bit of positive in the American system. I don't support the removal of abortion law, nor do I like mixing religion with politics. But, due to the system that exists in the USA, even though the government removed the right to abortion, the individual states can still support it and form their own laws around it. However, a similar situation in other countries would result in complet…

Give it like a year, I promise you it will be banned federally

A bill banning abortion nationally would need 60 votes in the Senate, which it will not get.

They already started this work yesterday: https://www.politico.com/live-updates/2025/01/22/congress/se...

Re: Trae: An AI-powered IDE by ByteDance

#364

Post by nuancebydefault »

elicksaur wrote: post #360
ioulaum wrote: post #289

Earlier quoted context omitted.

If it improves your performance by 1%, then for a salary of $2000/mo, that $20/mo is breakeven. If it benefits you more than 1%, then you're in profit. Of course, if you're in a job that doesn't actually care about performance, and performing won't lead to better salary at some point, then it may not matter.

Only if it increases your compensation by that 1%? I get making economic/stats based analysis like this, but is your boss going to notice that 1% to give you a raise they otherwise wouldn’t? Probably not… Your company culture can be performance-minded and this still be true.

What a weird way of thinking. If you act so pedantic anyway, you can stop coding simply 1 percent earlier each day, since your boss would not notice anyway. Personally I live in Belgium and I would be glad to pay 20 euros per month if it increases my productivity, as noticed by my boss or not.

The real problem however is, I cannot simply share my employer's repos to be absorbed by any LLM out there. So I use only the tools that my employer provides and approves of. Currently that is Microsoft Copilot chat/RAG via my work account. It takes some copy/paste and adaptations of problems/solutions but it is much more efficient than using SO. It is also a great teacher that never gets tired of my plenty why/how questions.

In my view, the future is that LLMs can train on entire private code repos until it understands its ins and outs. Currently it would need to fit in the context window, hence you need to babyspoon it, as I understand things.

Re: Trae: An AI-powered IDE by ByteDance

#365

Post by nuancebydefault »

knowitnone wrote: post #345
littlestymaar wrote: post #186

Earlier quoted context omitted.

I'm glad people care about their personal data, I just wished they cared as much when talking about ChatGPT, because in reality if you're not a Chinese dissident living abroad, then OpenAI is much more dangerous to you than the CCP.

OpenAI does not have a history of disappearing people and when they return, their attitude has completely changed (see Jack Ma). OpenAI does not have a history of aggression against other nations shipping vessels. OpenAI does not have a history of corporate espionage and IP theft. OpenAI does not have a history of setting up police stations in other nations. OpenAI does not have a history of operating an illegal bio-…

About the virus: you really need to be less in the company of people thinking up conspiracy theories. Remember that a pandemic has always been predicted in a 'not if but when' sense.

For the rest of the provided history feats, I can't really tell but most of it appeared indeed in established newschannels.

Re: Trae: An AI-powered IDE by ByteDance

#366

Post by nuancebydefault »

latexr wrote: post #236
animuchan wrote: post #230

Earlier quoted context omitted.

I'm thinking about this the same way, if a bit more selfishly: as a non-US citizen, I might want to move to the US in the future. Which is why I would like the US govt to have as little data as possible about me (who knows what the mainstream politics will look like, things that are very innocent today might be punishable by death in 10 years). Chinese govt, on the other hand, can have my data freely, since I don't t…

> I might want to move to the US in the future. > (…) (who knows what the mainstream politics will look like, things that are very innocent today might be punishable by death in 10 years). Maybe it’s just me, but I don’t think I’d ever consider moving to a country where I believed politics to be so unstable that an action could go from “very innocent” to “punishable by death” in the span of a decade.

Maybe it's just me but... personally I wouldn't move to a country in which more than half of the voters are pro a public jury-convicted criminal. I'm also reluctant to move to any country in which proper healthcare is reserved for the rich. I can't imagine moving to a country that is pro-life and pro-guns at the same time. A country that approves of, let alone encourages, getting rich by pump and dump of virtual freshly invented coins. Same sentiment i have about a country that approves of deportation of illegal immigrants, poor or rich, child or adult, criminal offender or not.

Re: Trae: An AI-powered IDE by ByteDance

#367

Post by spencerflem »

redcobra762 wrote: post #363
spencerflem wrote: post #327

Earlier quoted context omitted.

Give it like a year, I promise you it will be banned federally

A bill banning abortion nationally would need 60 votes in the Senate, which it will not get. They already started this work yesterday: https://www.politico.com/live-updates/2025/01/22/congress/se...

They're persistent - I wouldn't be surprised if some senators are found to be unamerican and impeached

Re: Trae: An AI-powered IDE by ByteDance

#368

Post by Aloisius »

ioulaum wrote: post #285
nikkwong wrote: post #154

Earlier quoted context omitted.

Absolutely and verifiably false, again! There are numerous data points that prove that the CCP exploits the TikTok algorithm to influence western views towards a pro-eastern bend. See my comment history for references. The close-doored national intelligence briefing that the senate got about the way the CCP manipulates the algorithm and on how TikTok is a national security threat resulted in a 100-0 vote in favor of…

They have all the US data on Oracle servers, with Oracle having the responsibility to monitor all access and make sure that no Chinese entity can get access. All algorithms are monitored by Oracle such that only user actions can determine what they see. So, unless you're claiming that "Oracle" is actively colluding with China, you're letting your worldview be defined by government propaganda. The Pro-Palestine thing…

We all know that's b.s. after the leaked audio recordings from TikTok meetings describing how US user data is still accessed from China.

> "Everything is seen in China," said a member of TikTok’s Trust and Safety department in a September 2021 meeting. In another September meeting, a director referred to one Beijing-based engineer as a "Master Admin" who "has access to everything."

The idea that Oracle can prevent Chinese employees from accessing data when the servers are running code written by ByteDance on bare metal and the US team literally reports to the Chinese one is, frankly, ridiculous.

https://www.buzzfeednews.com/article/emilybakerwhite/tiktok-...

Re: Trae: An AI-powered IDE by ByteDance

#369

Post by Coryodaniel »

diggan wrote: post #311
doix wrote: post #181

I see many comments about not using this because China/ByteDance stealing data, etc. I get it for for enterprise, you don't want your code leaking to the outside world, even if in my experience, the code is mostly garbage and not worth that much. But if it's for some personal project that you're putting on Github, does it matter? If my code is going on Github anyway, it's going to get slurped up regardless. I don't p…

> But if it's for some personal project that you're putting on Github, does it matter? I don't think the fear is that they'll steal code you'll end putting publicly on GitHub, but everything else. I guess there is some fear that it won't just analyze and process what you currently have open, but might scrape your computer for more data and so on. I personally don't believe ByteDance would be stupid enough to even att…

Trust em or not … had that happen to someone else using Cursor a few weeks back. I ran a prompt for refactoring a function and all of a sudden it started dumping someone else’s directory structure and file contents into my text file. Like hundreds of files. Not exactly sure what happened but I immediately uninstalled Cursor.

Re: Trae: An AI-powered IDE by ByteDance

#370

Post by guappa »

redcobra762 wrote: post #297
guappa wrote: post #264

Earlier quoted context omitted.

I don't see tanks in taiwan… What I see is this: https://en.wikipedia.org/wiki/List_of_countries_with_oversea... As I've already said, I never claimed china are the good guys. Just that they aren't the worse guys.

...you realize the US is invited to be in those countries, right? And why would you see enemy tanks on an island? What you do see are many many Chinese ships and planes, constantly violating Taiwanese air and sea boundaries.

"invited" :D
Post reply on HN