HN user
P and NP are classes of computational problems.
A problem in P can be solved in polynomial time - the computation required grows relatively slowly as the input size increases. Like sorting a list of numbers.
A problem in NP requires exponential time or greater, but a proposed solution can be verified quickly. For example, checking a completed Sudoku puzzle.
It is believed but unproven that all problems in NP are NOT in P.
The danger of anthropomorphism is not we elevate the machines, it's that we debase humanity.
I also think different ideas get conflated. It may be possible to build a machine that is super-human in the sense it can outperform the human brain in all kinds of measurable ways. Does not imply it possesses all the same qualities of the brain.
I respect a number of things Anthropic has published about the ethical issues at stake. But, having an in-house philosopher does invite you to make all kinds of unfalsifiable claims.
Everybody always wants to talk about job losses. It's only part of the larger imperative of preserving human dignity.
We don't really have leaders with the maturity and perspective (and lack of self interest) certainly in Government and questionably in tech that can be trusted to advocate for human dignity, so the release of this document from the Pope is a remarkable event.
Hello. Interesting project! Haven't gone through it yet, but want to consider using this in my CS master's capstone. While you have benchmarks I may create my own specific scenarios and comparisons vis-a-vis hosted inference to highlight specific economic benefit. Any suggestions?
Don’t quite agree with the implication that if the answer to a problem is readily available (from an LLM) there is no use in the struggle to find the solution.
However I think it’s very important to approach such questions objectively, or at least self uninterested, and not as one who’s worried about one’s job or sense of self worth threatened by LLM technology.
The value is in the development of one’s own mental faculties. In math classes they tell you you have to work the problems. Even if LLMs become capable of solving entire classes of problems that that set expands over time, the value in developing one’s ability never goes out of style.
Ever consider there's reasons to study Computer Science at the collegiate level, other than making yourself a more desirable worker?
However if you view your content as valuable and the algorithm does not anymore, it's probably not the best platform for you to be on.
Haha. Was just thinking that as I was reading a comment!
I was thinking, this argument is suspicously cogent!
Have had good results on MacOS just using codex (or your cli of choice).
Have it create a swift app, unless extended permissions are needed it can compile withouy going into xcode.
Few simple util apps, disk cleaner, clipboard manager. Worked pretty well.
Had better results than using xcode's built in ai extension.
I wouldn't underestimate this as a good business decision either.
When the mass surveillance scandal, or first time a building with 100 innocent people get destroyed by autonomous AI, the company that built is gonna get blamed.
Gorgeous UI! Will try it out.
I know about "the entire developer world has been refactored" and all, but what exactly does this thing do?
Runs git checkpoint every time an agent makes changes?
Made a quick bot app (OC clone). For me I just want to iMessage it - but do not want to give Full Disk rights to terminal (to read the imessage db).
Uses Mlx for local llm on apple silicon. Performance has been pretty good for a basic spec M4 mini.
Nor install the little apps that I don't know what they're doing and reading my chat history and mac system folders.
What I did was create a shortcut on my iphone to write imessages to an iCloud file, which syncs to my mac mini (quick) - and the script loop on the mini to process my messages. It works.
Wonder if others have ideas so I can iMessage the bot, im in iMessage and don't really want to use another app.
Compilation is transforming one computing model to another. LLMs aren't great at everything, but seem particularly well suited for this purpose.
One of the first things I tried to have an llm do is transpile. These days that works really well. You find an interesting project in python, i'm a js guy, boom js version. Very helpful.
Useless if it takes > 5 sec. to load a page
I disagree I think we always need new languages. Every language over time becomes more and more unnecessarily complex.
It's just part of the software lifecycle. People think their job is to "write code" and that means everything becomes more and more features, more abstractions, more complex, more "five different ways to do one thing".
Many many examples, C++, Java esp circa 2000-2010 and on and on and on. There's no hope for older languages. We need simpler languages.
We don't understand the brain. We fully understand what LLM are doing, humans built them. The idea we don't understand what LLMs are doing is magical. Magical is good for clicks and fundraising.
Been WFH for many years, but I can understand why some companies prefer (demand) in office. Sorry, but unless its written into employment law, no-one has a "right" to it.
I would think this could be a perk that companies could use to get an advantage in hiring. Although, maybe those who it appeals to on the whole may be lower performers?
Would love to see some real data on this.
I’d also distinguish between the hacker to gets access to a forbidden system out of curiosity or for a challenge, from a person who pays a ddos service to attack someone they don’t like (one of the accused actions of this kid).
The latter displays no competency in hacking or cybersecurity, only the attempt to harm another.
My concern in their access to secure government systems is not their hacking competency (which has not been demonstrated), but their sociopathy which has.
Maybe because nobody really asked for these things gemoji and "rewrite my email".
Neural nets are a foundational technology there should be a ecosystem of apps based on it. People pay 1500$ for an iphone because it has apps, not to call and text people. I think its coming and Apple can't just NOT do AI, so I get it, but it will take a little time.
If ya don't want to do includes,
` window.$ = document.querySelector.bind(document) window.$$ = document.querySelectorAll.bind(document) `
Facinating software although I hope the idea "we're gonna rely on people to be good humans and DO THE RIGHT THING" is quickly abandoned and instead there is just as robust development of detection software that goes along with newer and better deep fake tools.
3.5" disks, nice. We were still on 5 1/4" in high school in 87-88.. True floppys! I remember thinking 3.5" hard disk - wow the technology :)
Increasing enrollment means increasing faculty.
Maybe controversial but I really believe this lowers overall standards. You just can't find enough faculty to meet the high standards.
Look what happened when tech blew up - they hired like crazy. I'd argue the overall level of programming skill went down.
Always felt that the ability to think was indicative of academic success.
But also a good SAT for (those not naturally brilliant = most of us) also a positive indication of the student work ethic (to study for the test).
The problem I believe with the disadvantged students argument, which I believe to be a very real problem - is that lowering the standards wrt the test scores - is just a bad solution to a real problem.
Poor students do sit at a disadvantage - but the fix is much more difficult and complicated, and involves that child's entire life experience until they become 17 or 18 y/old. No one has a handle on how to fix it.
I get the feeling smart + hard work will come back into fashion when it comes to academic opportunities.
Testing and QA is great and all however there is something to be gained from a CompSci education besides making more money for the company. It can be enjoyed just for the sake of it.
Binary arithmetic doesn't make dollars for your bosses, but its fun. That doesn't make it less significant.
In 50+ years I've learned one thing.
When you have a kid. Drop EVERYTHING in your life, take off (completely) from work and spend your days with them. For at least a month or two. Make breakfast, watch little kid shows, all that.
Yeah money, but you have decades to earn money. Those couple months you won't get back.
Donors like that threaten to pull their money when you don’t behave. It’s why they donate. There is no such thing as a donation.
Have paid school tuitions for 15 years now, pre-K to 12, and in 2nd year of college now. Some thoughts..
College is just like private K-12 - there is no set price, it's indexed to your income. They have to set a high list price so the wealthy can pay that and they aid those w less income. Don't forget about car and health insurance!
Do every single aid and scholarship app you can find. Use the appeals process especially if you have extenuating circumstances. Read the rules carefully about which assets/income is relevant.
If you're divorced (I am), read the rules carefully about income and asset reporting in those circumstances.
It's not that you're paying for the credential, you're paying for the experience, education, and credential. Just the credential is the expensive part of it.
Example, I've been a DJ (just for fun) for 30 years. Just saw an ad for DJ school. Even if they wanted to be a professional DJ (ok, if that's your dream), I'd NEVER pay for my kid to do it. There's a legit cost in educating anyone. But the credential is worthless. A university degree has some value. Of course, you can learn ANYTHING online for free. Anything. The credential is the expensive part of it. So make sure it's a credential with value.
You may have to say NO to your kid(s). This is VERY tough. But weigh your options carefully as spending your entire retirement fund on college (mostly for the credential) is a dubious proposition.
Built some log databases and back end frameworks myself with some of the same concepts. I applaud the creativity in rethinking how back ends should work. Please now do frontends next! :)
"But it's not a fully functioning 2023 Twitter!!!" I think some people miss the point. This is not about hey we built a Twitter clone. This is about a POC for a novel app architecture.
We need to be constantly examining and re-examining our thoughts about the best way to deal with distributed systems, scale, developer workflow. Even inventing new ones.