I spent 8 years in academia (2004 - 2012) before moving to Industry. But as I've aged I've thinking of going back. I made it good in Industry so I have enough to jump to Academia without worrying about money... but I just hated the "publish or perish" mentality, and writing papers (I wonder what's the state of that now with LLMs... back in the day I was reviewer for some journals and most papers were pretty bad).
I understand your point and clearly see that LLMs cannot be compared to audio ... but ...
Back when I was a kid, music, audio and sound systems had high quality as a standard.
Nowadays people listen to music mostly with bluetooth headphones which basically recompress an already compressed audio signal to send them in low quality. Also, it is more and more difficult to find OK stereos that play music in good quality. Either, you have to pay very high prices for overpriced "audiophile" equipment, or you are stuck with cheap chinese MP3 players.
Yet, society and markets have spoken. Sometimes society is happy to accept marginally worse products in exchange of price and convenience.
This hit very close home. I'm a 44 year old developer, with Software Engineering Bachellors and CompSci MPhil and PhD. All my life I spearheaded "best practices" and code quality (from Fred Brooks, Joel Sposky, Martin Fowler, etc...).
But since LLMs arrived... things have become crazy. The layer of "obscurity" that permeates code writing seems to make a lot of those "standards" moot or just not really pragmatically possible to follow.
Breaking the export rules. Tech workers should be used to the idea of a "Invention Assignment Agreement".
Manus was built in China and all of its development happened there. In order to skirt Chinese review of the deal they tried to close down shop there and move to Singapore.
I don't think China is being unreasonable. I'm sure the US would act exactly the same way if an American tech company raised money from China and then tried to close down in the US and move all of its IP and technology to a different country so that it can be bought out by Alibaba or Bytedance without having to deal with US approval
There is no equivalent exit ban in the US that can be instituted on a whim for regulatory or business disputes. If you want to know more, you can read up on it in this Stanford Journal of International Law publication:
Do you read the news? Whether or not to stop Nippon Steel's acquisition of U.S. Steel was being discussed everywhere. On what basis was that power?
> Nippon Steel's acquisition of U.S. Steel can be stopped by the US President based on a recommendation from the Committee on Foreign Investment in the United States (CFIUS), citing risks to national security under Section 721 of the Defense Production Act.
National security risks. Exactly what China is citing. It's literally the exact same situation.
> Two days ago, President Trump issued an order blocking the $1.3 billion sale of a Portland, Ore.-based company called Lattice Semiconductor to private equity firm Canyon Bridge Capital Partners. The stated rationale for Trump’s order was national security.
Outside of immigration issues, you can only be made to surrender your passport if you have been arrested and indicted for a crime, as a part of bail. That power can only be granted by a judge.
China arbitrarily traps people in China without any such thing or any due process whatsoever.
The first case makes sense: ex-CIA officer explicitly outing CIA officers. Naturally, the government is going to step in and it's a false equivalence to compare to restricting random citizens.
As for your second case, US schools teach about the perils of McCarthyism. You neglected to link to the subsequent Supreme Court ruling in 1958 overturning the confiscation of the passport over protected speech. Note how long ago that was and how it's taught as a black stain on US history.
Anyone with a child support order that makes decent money is only one misrecorded or bounced payment away from being ineligible for a passport. The trigger is only 4 digits of USD.
So were these founders found violating a child support order?I'm still unclear on what crime they actually committed and what they're being investigated for.
Is it possible that they are merely pawns in a political dispute between two rival countries?
It's one thing to block an acquisition because you don't want your rival to gain an advantage, an action which is not limited to the CCP.
It's another thing to detain an individual when no crime was committed.
"Missing" (but quite often only due to a clerical misreporting) a payment isn't facially criminal and isn't even established as "violation" without a contempt hearing where you can argue why you didn't actually violate it. So the passport denial is even looser than that.
I'm just pointing out the bar isn't much different except dressed up in a think of the children meme. I'm not justifying either one.
Understood. I'd note that the difference is that with a missed payment, you can simply make that payment (which can be a relatively small one as you noted) and you're free and clear.
With these Chinese founders, I'm not sure it's quite so simple.
In the US, the Passport Denial Program, since 1998 (other developed countries enacted similar legislation), following the 1992 UN Convention on the Rights of the Child (UNCRC) [2]:
> The Child Support Enforcement Passport Denial Program was enacted as part of the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. While authorized in 1996, the program was jointly implemented by the U.S. Department of State and the Federal Office of Child Support Enforcement in June 1998.
> China arbitrarily traps people in China without any such thing or any due process whatsoever.
What makes you think there's no legal process for blocking nationals from leaving China?It's a very common instrument and in a bunch of countries it's an administrative measure with even less scrutinity than a judicial mandate. Do you consider France or the UK to be a countries without rule of law or due process?
But to the point in the US, for example, the government can just issue a warrant for you as a material witness or flag your passport and then you can't leave; these are hardly due processes and more like legal workarounds to do exactly the same thing; the US has disappeared plenty of people in much more sinister ways than that, however, so I agree that there's no equivalence here: the US is worse.
America is not exactly a shining moral example for the world, particularly these days, but these Chinese apologist takes can be a bit baffling to read at times.
It mostly doesn't make any sense and seems to be motivated by some kind of animus or bigotry. But maybe understandable given the current administration's behavior.
Funny, we now enter the era of "Made with Handcrafted Code" or "Handmade" . Same way as furniture, carpets and any other "handcrafts" are made now... or Lamborghinis
I want to focus on the "colleagues submit thousands of AI generated lines of code for review" comment.
Humanity developed Code and programming languages for people. They are supposed to provide sufficient expressiveness so that we people can understand what is happening, and 0 ambiguity, so that the machine can perform is instructions.
But computer code has been a way to communicate among us people on our intentions (what we intend the machine to do). Otherwise, we would still be writing in assembler.
But now, computers are generating code, A LOT of code. So much, that it's becoming more and more difficult to stay on top with our verbose languages.
We will need to develop a better way for the computers to a) produce the instructions to perform the tasks we tell them to , b) produce reports or some accessible way for us people to understand and share what the instructions are doing.
I'm wondering whether the layoffs are partly targeting people who haven't adapted to using AI tools, particularly those who are openly dismissive of AI-assisted work.
Because the job itself has now changed, and they haven't. Their output speed might have been eclipsed by that of the engineers who efficiently adopted the new tooling.
Where I work, the power dynamics have shifted wildly. There are a number of senior engineers who refuse to touch the stuff, and as a result, they can barely keep up with their peers. Some of our juniors are now running laps around them.
When a stranger to your craft can now teach themselves what you know, how to do your job, and even how to automate your tasks in the span of the same workday as you, all while reliably being able to gauge the innacuracy of the output they're reading, how much longer do you really hold relevance?
Because the job changed out from under them - it's now to use AI as much as possible and generate so much and so convoluted content that humans have no chance of keeping up the "velocity" without being entirely dependent on it.
So, yeah, the US is no "blanca palomita" at all. And those of us suffering from their actions have learned that all powerful nations have good and bad things. Here in Mexico, we've got BYD cars, and they are AMAZING. Also being able to use DeepSeek is so cool.
If your government refuses to stop the flow of drugs into the US by addressing cartels don't be surprised if the US delivers weapons to said cartels so they can have some infighting going on.
If the mexican government would actually make work of dismantling the organized trade, there would be no incentive to deliver them weapons to shoot each other.
supply is never an issue, USA would supply poison to entire planet if the demand was there. blaming Mexico for the sickness of our society is very rich (but often repeated)
This really should be the case. If AI tools are really making it easier to build stuff, we should see hordes of new startups solving all kinds of problems thar were difficult or expensive to solve before.
I've been seeing this in the startup ive been for the past year. We are 20 people, and are solving fiscal reconciliation problems for HUGE companies in my country. Building thing that were just not scalable before.
I'm waiting for all the cool startups in both b2b and b2c that solve health, time spending or money problems.
in theory yes, but all the money is going into AI or AI-adjacent startups that no one would actually build a product that solves problems if it doesn't incorporate AI in it.
Using AI is fine. The key is to use it to build processes/Systems that solve problems deterministically. Instead of "asking them" to solve the problems non-deterministically themselves. It's way cheaper and robust.
As an example, we had to be able to parse most uses b2b bank statements. That means understanding the structure of around 30 different formats, some with very subtle differences within them.
The naive and expensive approach was to train an LLM to do it (after OCR).
Instead we used AI to generate a generalized python parser that is "configurable" for different structures. (And also extracts data from PDFs without OCR, unless they are pure images).
It covers the 99% of the cases. And for that 1% we pass it through AI for immediate solution, and to generate the additional deterministic config to cover it.
But I digress (lol). The point is, there are so many interesting problems that can now be solved. We should have teams of 3 to 5 people doing crazy stuff.
I spent 8 years in academia (2004 - 2012) before moving to Industry. But as I've aged I've thinking of going back. I made it good in Industry so I have enough to jump to Academia without worrying about money... but I just hated the "publish or perish" mentality, and writing papers (I wonder what's the state of that now with LLMs... back in the day I was reviewer for some journals and most papers were pretty bad).
reply