hodder

Everyone is so convinced that AI is bubble, it is hilarious. This market doesn't look extended to me given the fundamentals.

Exactly. Probably an hour if you just tell your model of choice to do it for you and implement a logical testing framework.

Honestly migrating back is trivial now. It just isn't that big of a pain.

Nonsense. Plenty of VERY good software that has been built at immense speed. Look at Muse as an example.

Agreed, but it is also much easier to get something made in the first place. LLMs enable rapid prototyping and rapid shifting to solve real problems. My applications are morphing from mediocre to highly useful problem solving machines much more rapidly now.

Fascincating article. Also a cool band name.

It misses the single most powerful positive though. It is incredibly fast and empowering to the point that putting up with the four horsemen is easily justified. Coding by hand is largely in the rearview mirror unless you are more interested in the craft of writing code than simply using code to get things done.

Highly unlikely an entire chain of military personnel knew it was a school and was like, "yeah get em good, and lets see when they come scrambling in for their kids corpses".

Do you hear yourself? Have you ever in your life met someone this sadistic? It was definitely a mistake. One might argue criminal negligence, but there is no way everyone in the loop here is just pure evil. That just isn't how people think in North America. It is unfathomable that a team of US military (people with families of their own no doubt), would all nod along and agree with your contention.

This is a solution looking for a problem, built to satisfy the creator instead of serving the buyer. This is exactly what PG warns against doing. One would think Sam would know this too.

Amazing. Best post I've seen on HN in ages.

The thing is, Apple will have never lost this opportunity. The distribution advantage of the user base is so deep they can basically always implement this at will.

Cosmic expansion happens uniformly rather than through a local gravitational pull.

In physics, "we" use the raisin bread analogy where when you bake the bread, the raisins are all getting uniformly distant from each other with no central point of expansion and no obvious sign of pulling in any specific direction such as would be the case for a gravitational effect.

The problem is no one wants a META agent. so yes LLMs are great. And still no one wants to use a Zuckerberg one.

muse is probably an awesome technological innovation and will still completely flop bc it’s a consumer product outside of social from Meta. It’s a solution in search of a problem. I get why the company wants this. I don’t get why any consumer does.

I don’t get how this differs from any other optimization harness. The problem is clear, but I’m not sure on what this company is doing that is unique to solve it.

You mean like Carney did in January? Trump cannot be shown reason once he has convinced himself his gut feel is correct.

The ideal VC startups are low capex, high ROIC companies.

Energy almost universally requires massive capex to achieve a dime in revenue, and capital intensity never really subsides. YC has funded energy focused startups in the past (eg. Oklo and Helion), but it certainly cannot compete with the kind of scale and ROIC you can achieve in the digital world. Energy is difficult for seed investors.

Id be more weary of saying I didn't use AI.

People repeat this rhetoric all the time, but I promise you, finance professionals can and do easily migrate to other messaging platforms such as ICE chat. This isn't a moat for BBG

Excel succeeds in the business world because users interact with visible data directly on the screen rather than relying on abstract database storage and retrieval.

Non-technical workers can build functional tools rapidly, outperforming traditional software development speeds massively. Engineering CS types completely miss the point on this. If Im making a model for finance I need to see the data live, update fast, get my charts done without messing about with matplotlib etc. Get to solutions and decisions.

Spreadsheets are still prolific bc they enable rapid prototyping, visibility and flexibility that is to this day completely unmatched.

And not a single person cares. What does fully valid even mean anymore?

This is like the stupid Meta Portal. It doesnt matter how good it is, no one will use it bc no one trusts them. Right or wrong.

Economically if you are vastly outcompeted by other programmers on productivity, yes that is "no longer of value" from a purely employment perspective. Much like an old person who cant use a computer has little value in the job market beyond being a greeter at Walmart, a programmer who hand codes a loop is next to useless on a productivity basis such that it makes zero sense to employ them. It is unfortunate but true. Why pay someone to accomplish less per dollar pf cost. Feels?

Sure. But the market is littered with companies you can invest in. You don't "need" to bet on any single one of them.

But why would you want to buy a company where the valuation reflects an extremely optimistic outcome already? Basically you are assuming obscene growth from SpaceX in AI, datacenters, rocket launches etc just to for the stock to go nowhere. The market cap was over 2T!

When you buy equities, if the company turns out shockingly successful beyond most peoples probable expectations you dont simply want your investment to just stay flat, you want to make multiples of your money. In order to do this, you simply cant start from a 2T marketcap. No company worth 2T can make you 100x your money or even 10x or 5x your money in a reasonable amount of time given the overall size of the economy.

I agree man. I’m just quoting Donald.

Forgive me because I do not understand the supply chain for memory. With Micron et al effectively scalping their customers with an oligopoly on probably the lowest intellectual IP in the chain, does this not guarantee 10 years from now a) We are either overbuilt as hyperscalers cut capex, or b) hyperscalers vertically integrate. Or is it truly that hard to make memory? And if that is not true, perhaps it isn't really a commodity at all.

Honestly Jassey, Zuck and Tim Apple are prob on the phone with Donnie. If oil companies are “gouging,” what is 85% margins on memory, threatening the whole bull run and raising compute, Killing AI, and raising iPhone/computer pricing? Countdown to DOJ antitrust case is ticking.

To be clear: I understand how markets work, Im just quoting Donald Trump's tweet from yesterday calling oil companies gouging, and I predict government intervention and polital pressures regardless of economic realities.

Forgive me because I do not understand the supply chain for memory. With Micron et al effectively scalping their customers with an oligopoly on probably the lowest intellectual IP in the chain, does this not guarantee 10 years from now a) We are either overbuilt as hyperscalers cut capex, or b) hyperscalers vertically integrate. Or is it truly that hard to make memory?

And if that is not true, perhaps it isn't really a commodity at all.

The claim that Alberta is actively trying to get rid of all wind and solar development is internet hyperbole that ignores real capacity data. Alberta actually ranks second in Canada for clean energy growth, and its renewable output surged by over 25% year-over-year into 2026.

The high-profile project cancellations people point to weren't a government ban. They happened because the province changed its transmission rules. Previously, ratepayers subsidized the massive utility costs required to connect remote wind and solar farms to the central grid. The province ended this, forcing private developers to internalize their own grid connection costs. Once forced to pay for their own infrastructure, highly speculative, unfinanced projects simply became economically unviable and dropped out of the queue.

If a private wind or solar developer wanted to build a massive farm in a remote, rural area (like Southern Alberta) where land is cheap but high-voltage power lines do not exist, they only had to pay for the immediate wire connecting their project to the nearest local substation. Taxpayers were subsidizing those players, because it was a "load pays" system.

Please do not fall pray to the general trope that Alberta is a backwards hillbilly province. Subsidizing private developments with public money is not something that should be encouraged.

On Canada broadly, you are correct in your baseload numbers and I agree with you.

(Energy trader here)

Your arguments are much lower quality than an LLM will generate. You are so transparently biased against them it is shocking. I’d certainly rather learn from an LLM than someone like you who is both grossly misinformed and a dick about it.