HN user
You are more proficient at writing code than ai specs.
You could get better/faster at writing good specs and its a higher cap skill now.
For low risk changes you could let llm write its own spec from high level requirements and just validate its assumptions/design decisions.
But that's mostly just vanilla models.
Some critical parts of agentic development are literally months old while most mature ones are merely a couple years old.
Critical thinking was the most critical checkpoint for models to reach.
Now its only a matter of time for the right tooling and orchestration to come along. Once agent memory and recall works well, all of those bottlenecks will go away one by one.
I think everyone understands models will be a commodity.
Its the user base (with ads and upselling) and proprietary wrappers which will make money for typical customer.
Even enterprise customers arent going to be spending a lot on tokens. Once labs no longer have to subsidize trainings tokens costs will drop 10x and once models get burned on chips costs will drop 10x more and you physically won't be able to burn significant number of tokens unless you're deliberately trying to.
If startups includes openclaw users then I could see this being true.
Deepseek is barely behind frontier models while 10x cheaper and 99% discount for cache.
I think deliberate practice is what's really core to improving any skill, including memory.
Spaced repetition is an effective way to review things but its biggest benefit is a process that's easy to be consistent with.
Somebody else can have equal or better performance with other technique but just like dieting, it doesnt matter as much what method you use as long as you stick with it.
Starvation is already not a thing anymore.
Sounds dumb, but then also if something unexpectedly had a list of 20 ingredients you might reconsider buying it
28% more visits
So, from 3 to 4 people?
if you feed it back any of the claims, it defends them pretty well
it doesn't deny any of the stats and cites none of it is secret as it's all published on the public dashboard
at the same time, churn is expected for such a company, doesn't mean it's not a real business, especially as AI gets better and margins improve... I think their business model is real
probably worth 30MM too since market is untapped, cost of user acquisition is much less than average profit per user and if one of the businesses actually takes off, they have 100% vendor lock-in and take 20% off the top
You have all that with Gacha pulls and battle system.
It certainly doesn't seem like consciousness exists. Although to disprove that hypothesis all we need is to find a single counter-example, which coincidentally all of us can provide via our personal experience of self.
It would be fine for an unconsciouss intelligence to maintain that hypothesis lacking any evidence to the contrary, but for us it seems we are just all gaslighting ourselves to ignore the one counter example we all have.
It's a bad example, because both sides actually got the entertainment they paid for and is totally valid economic activity.
I'm arguing it's not "success". I don't believe that a meme stock is a truly long-term business.
Duh. Crypto is the peak meme investment. Literally worthless, yet the market size is in trillions now, with even pension funds buying into it.
It doesn't follow any fundamentals, but as the meme markets defacto exist, there needs to be some models of valuing these investments other than fundamentals.
It seems like the incentives are to think short term and do sketchy shit to pump the stock.
Its a meme stock, that's what stockholders actually want.
Depends on what they're trying to incentivise.
It's quite possible they aren't trying to measure performance but are literally just trying to increase token consumption to feed the bubble and hype.
Plus pressure employees may find new unique use cases for AI.
It's like if your goal is inflation, you give out tons of money and as long as its spent, you achieve your goal.
Doesn't matter as long as incentives are aligned. As a major stockholder and active advocate, he is already vested in success.
Based on how same models rank fluctuates week to week, all I can conclude is that no frontier models is statistically better than the other or it's too task dependent that the result cannot converge.
I think kids and their free labor were the biggest wealth generating asset for the poor and as such wouldn't be given away except in the most extreme circumstances.
Planetary Annihilation did it and wrote and gave talks about it.
The players barely ever change. People don't have problems following sports, you shouldn't struggle so much with this once you accept top spot changes.
That's way more than 10, around 50
legislation dampens this behavior
Potentially. Well meaning and thought out legislation still distorts the markets, possibly making things objectively worse.
Capitalists claim that this is optimal.
It's more optimal than planned economies until we have AI planned economies with realtime feedback, I guess.
Consumers get cheap goods during oversupply and most inefficient companies get elliminated during bust while consolidation leads to economies of scale.
That's very interesting, thank you!
How does this compare to Jules from Google?
Do you think it's just part of their training set now?
They are not doing random rotation, simplification here means they are aligning the outliers. If you threw a bunch of shapes on the ground they are picking up one that rolled away and putting it with the others.
How can a boolean value preserve all of the relational and positional information between data points?
They aren't reducing entire vector to a bollean only each of its dimensions.
AI capability problem is mostly solved; the distribution and trust problem isn't.
SaaS opportunity? Maybe, some sort of marketplace of AI-written applications and services with discovery features?
I have a junior position open and got 1,300 applicants in 1 week before we took it down. Many of the candidates with strong resumes are just lying and doing so well enough to pass HR screens.
I doubt any sort of AI screen would help though as many of the lying candidates are already using AI assist tools making it just a cat and mouse race...
I don't know a good solution to give everyone a fair chance.
I think you are agreeing with me since my point was that AI is a force multiplier and will enable manual artists to create (and iterate) faster and on much bigger projects.
Prices will go down, not because people can directly generate art, but because you could hire an artist to complete the same scope of work for 1/10th the cost.