gspetr
Really, compress. Compression is a useful proxy for intelligence.
To compress data efficiently, an entity must discover the underlying patterns, structures, and regularities within that data. Identifying these structures is precisely what we define as "understanding" or "intelligence.
Anything we put into writing can be thought of as a string of data. The Kolmogorov complexity of a string of data is the length of the shortest possible computer program that can output that string.
A good compressor doesn't just memorize (yes, there's a reason LLM writing features the "it's not just X, it's Y pattern"), it builds a generalized model to handle unseen, similar data. LLMs are trained by compressing the internet into weights, they use this to write new code.
When an LLM is trained, it compresses ...bytes of human text into a few ...bytes of neural network weights. When you ask it a question, it uses that compressed model to generate (decompress) a coherent response.
Consider the dichotomy:
LLMs display emergent reasoning, coding capabilities, and logic because they were forced to compress human language so tightly.
LLMs often suffer from "hallucinations" and struggle with novel logic puzzles because compression encourages mimicking statistical patterns rather than building a true causal model of the physical world.
An algorithm can achieve massive compression if it utilizes a highly complex, slow, or massive decompression engine. If a system takes 10 000 years to decompress a file, or requires an infinite amount of computing power (like Kolmogorov complexity), it cannot function as a practical intelligence. Real intelligence must operate within tight time and energy constraints.
> Maybe there's true brilliance out there
Brilliance is a trope. pg's 2009 "The Anatomy of Determination" deals with it: https://www.paulgraham.com/determination.html
In it, he argues that determination is the single most important predictor of success, surpassing even raw intelligence.
While popular culture loves the narrative of the "brilliant genius," Graham's experience backing hundreds of startups taught him that a highly intelligent person without drive quickly becomes ineffectual. Conversely, a profoundly determined individual can find pathways to success even in ordinary or unglamorous industries. Over time, determination actually starts to look like talent.
I'm going to commit a faux pas (a breach of HN etiquette) and give an AI summary in 3 bullet points because, well, the article deals with 3 points, so it's a good fit and I doubt I can compress it better than AI:
1. Willfulness. The desire to get your way regardless of obstacles or consequences.Being fearless about consequences and pushing through "hellacious" moments when others quit.
2. Discipline. The ability to force yourself to do what you planned to do. Being "hard on yourself" and turning off the background analysis in your mind that questions the plan.
3. Ambition. The grandness of your destination and what you want to achieve. Choosing the right target; if determination is a vector, ambition is the direction it points.
Of these 3, it's quite obvious at this point that LLMs are superior at 1) and 2), as LLM alignment is quite good, and it's ability to spin up sub-agents that can work for over 24 hrs is better than humans. Human limit is basically 1 hackathon's worth of cognitive work. Beyond that, health becomes a concern, and how many people can produce useful work under sleep deprivation? Codex can work for 24-ish hours, as evidenced by today's "Auto-research with codex: How I achieved a 232x Faster Kernel": https://news.ycombinator.com/item?id=49309549
So what's left for mere mortals? Apparently, nourishing your ambition:
"So here in sum is how determination seems to work: it consists of willfulness balanced with discipline, aimed by ambition. And fortunately at least two of these three qualities can be cultivated. You may be able to increase your strength of will somewhat; you can definitely learn self-discipline; and almost everyone is practically malnourished when it comes to ambition."
The irony here is that you are correct in the original sense of the word "decimate".
However, in my lifetime I've never seen someone both understand the original meaning of the word "decimate" and use it as the ancient Romans intended.
They do, which is why only 45 countries observe DST.
25 observe it partially.
And the rest, roughly ~125 countries do not.
Historically, ~140 countries did.
To put this into perspective, only about 1.2B people out of 8.3B observe it today. Which puts you into a very small <15% minority.
Because there are countries where national or state governments do.
Why pass on it now? There has never been a better time in history to deal with it than today: AI can oneshot most of what you've described.
A typical American semi-truck (18-wheeler) can carry between 40,000 and 45,000 pounds (20 to 22.5 tons) can travel only 160k miles (2/3 distance to the Moon) on a budget of $400k (i.e. $20/kg).
Do you expect us to believe that rocket transportation would be in the ballpark of truck prices in the near future?
There's a famous quote attributed to the Italian military commander Gian Giacomo Trivulzio in 1499.
When asked by King Louis XII of France what preparations were needed to invade the Duchy of Milan, Trivulzio responded: "To carry out war, three things are necessary: money, money, and yet more money."
In this one there are often thousands of applicants for a position, most of which can't pass FizzBuzz.
Why would they (or anyone for that matter) choose a candidate who can't figure out how to find info about a trivial method?
The reason is the intent behind their question, which they don't vocalize.
This question means we are dealing with an extremely broad hiring funnel designed to fail people who can't FizzBuzz and need to keep answers at MVP level.
In other words, if you are asked to put out a fire use a bucket of sand, not a state-of-the-art fire extinguisher.
str.find(sub[, start[, end]])
"Return the lowest index in the string where substring sub is found within the slice s[start:end]. Optional arguments start and end are interpreted as in slice notation. Return -1 if sub is not found."
Your instinct to resort to "in" is correc,t as it's generally slower than the "in" membership test, but the interviewer has even allowed the use of Google. Blanking out after that is really bad.
The interviewer is not asking to solve a problem here, they're asking for a simple ability to follow instructions, hence the offer to use Google to find the correct answer.
You could make a very solid case for using "in" (it is 2-4x faster), but only after you've solved the task at hand, this is what is expected in interviews. Not knowing the interview meta makes an average Joe basically unhireable in this market.
Step 2: Parse the output with your eyes. The method is literally called "find".
This one-trick pony failure mode could perhaps have been fine for a guy who did Java and nothing but Java for 10 years, but you are supposedly the person who runs "pythonforengineers" website...
100% correct call by the interviewer.
Interestingly, LLMs disagree with you.
Your statement is only accurate in an extremely narrow case, like if you were there to hear the person speaking, before their speech which was transcribed. Obviously, it is not true for almost all of human writing.
And if you were to go commaless, you will quickly get to rather precarious sentences, such as this one:
"Let's eat grandma."
A comma is the natural fix:
"Let's eat, grandma."
sha256 is deterministic, LLMs are not, even at temperature set to 0.
1)Workers get more productive thanks to AI.
2)Higher worker productivity translates into lower prices.
3)Most importantly, consumer demand needs to explode in reaction to lower prices. And we're finding out in real-time that the demand is inelastic.
Around 1900, 40% of American workers worked in agriculture. Today, it's < 2%.
Which is similar to what we see with coding: The increase in demand has not exploded enough to offset the job-killing of each farmer being able to produce more food.
The same thing it's always been: The military-industrial complex.
> MAGA was sure it was backroom democrats.
It's not that hard to distinguish "them", just look at how fast the mainstream media threw Biden under the bus over Afghanistan withdrawal.
1)POTUS orders the withdrawal.
2)Generals botch the withdrawal on purpose.
3)Mainstream media (left and right) eviscerates the POTUS. This sends a strong message to this POTUS, as well as any subsequent Presidents: "Don't mess with the profits of the complex or else."
This was the tipping point for me when I realized that the deep state is not a just a bogeyman conjured up by the right wingers. Should you cross the complex, it will just as easily come for you even if you're a Democrat that's been in politics for 50 years.
Finally, the Atlantic is as establishment as it gets. No matter which party is in power, their editorial board serves the ruling class, of which almost nobody on HN is a part of.
Whether their interests align with yours or not you can ascertain just by looking at approval ratings of the US Congress.
Ben Horowitz recalled asking Zuck what his engineer onboarding process was when the latter complained to him about how it took them very long to make changes to code. He basically didn't have any.
[0] He holds the title of Chief Engineer at SpaceX.
Go ahead, try to delete the useless Microsoft Edge browser if you're not in a select few EU countries.
In my experience, you can't do it cleanly. Asking LLMs will tell you the following:
1)Modify a certain registry key to enable deletion. Which I did, but the only thing that accomplished is un-gray the delete button in the Control Panel. Once you press it nothing happens.
2)Windows will eventually reinstall Edge. So you're basically screwed.
B) makes things worse in several ways, but primarily by stifling innovation. Only large incumbents will have no trouble paying for the measures required to ensure compliance.
There's also the cost of enforcement, which will likely have to be borne by the taxpayers. I don't think this is a good thing to spend money on.
C) cannot be enforced, and any good faith attempts will cost more than the damage from harm they're supposed to prevent.
It's in a business' best interest to maximize demand for its products. Which is mostly fine for society, country, and the world by large if you're selling paper cups.
However, if you want to sell more weapons you are interested in lobbying for events that increase the consumption of weapons, in other words: wars.
See the problem yet?
It's not "people". One half of all people grows up playing contact sports or at least have some form of rough-and-tumble with their homies in schoolyards. This half also knows that you can get punched if things get too rowdy.
The other does not. Almost all of the entitled road blockers are in this category.
When Glitch failed in 2012, founder Stewart Butterfield offered to return the remaining $6 million to investors. Ben Horowitz instead encouraged Butterfield to pivot and build out the internal communication tool the team had developed for themselves, which eventually became Slack.
I saw an interview (don't have the link at hand unfortunately) where Horowitz said he didn't much care for the $6M as he had already been set at that point moneywise, and essentially wanted to gamble on an off chance Slack succeeds.
Horowitz continued to support the company through its rapid growth and eventual direct public offering (DPO) in 2019.
As was foretold in the Tyler Cowen's eponymous 2013 book "Average Is Over".
In it he argued that the modern economy will undergo a permanent shift where "average" performance no longer guarantees a stable, middle-class life.
He predicted that the economy will split into two distinct classes: a high-earning elite (roughly 10–15% of the population) who thrive by collaborating with technology, and a larger group (85–90%) facing stagnant wages and fewer opportunities.
AI summary of the other key points of that book:
The "Man + Machine" Advantage: Success will belong to those who can effectively use smart machines. Cowen uses Freestyle Chess (teams of humans and computers) as an analogy, noting that human intuition combined with machine processing power consistently outperforms either working alone.
The Power of Conscientiousness: In a world of abundant information, the scarcest and most valuable traits will be self-motivation, discipline, and the ability to focus. Hyper-Meritocracy: Advanced data and machine intelligence make it easier for employers to measure an individual's exact economic value. This leads to extreme salary inequality as top performers are identified and rewarded more precisely.
A New Social Contract: Cowen predicts a future where individuals must be more self-reliant. He suggests society will move toward lower-cost living models for the non-elite, featuring cheaper housing and "bread and circuses" in the form of low-cost digital entertainment and online education.
EDIT: Notice how we're basically already here: Netflix is cheap, YT is free, Khan Academy and MIT OCW is free, Coursera/Udemy/etc. are cheap.
Stagnant vs. Dynamic Sectors: The economic divide is worsened by "low accountability" sectors like education and healthcare, where productivity is hard to measure and costs continue to rise, unlike tech-driven sectors that see rapid gains.