HNHacker News
TopNewBestAskShowJobs

nitwit005

9,454 karma · joined December 9, 2014

submissionscomments
nitwit005··on MCP was always a bad idea?
Sure, but if you have a staff of 10000, you're doing this 10000 times. Unless you share that in a common place, and now you're re-inventing the thing we're claiming not to need.
nitwit005··on AI-generated posters don’t have to be horrible
I've seen two resdit threads where people insisted the hands were wrong, and people took photos of their hands in the same pose.

But, then, people did this for digital editing before. Insisted there were "clear signs of photoshop", from people who never opened that app.

nitwit005··on Microsoft exec called AI scraping 'the largest theft of labor in human history'
> Normally this isn't the case for any technology except for the time it first comes around.

Well, sure, there's no one left to fight once everyone already lost their job and moved to another career. But, that's sort of a "might makes right" resolution. The workers don't have the political sway needed to get the government to intervene.

Businesses succeed in getting that sort of market intervention all the time. Most of the modern changes to copyright law are driven by business lobbying.

nitwit005··on Most people prefer traditional architecture
Plenty of people will rate the design of classic cars more highly, and will visit shows that display them. They also won't buy a car that looks like that.
nitwit005··on Most people prefer traditional architecture
The people outside the building aren't paying rent.
nitwit005··on Java 27
Until quite recently, the reasonably feature complete open source libraries available for things like DICOM or HL7 were old C/C++ libraries, Java, and C#. That often created a choice between Java and C#. People not doing Windows based development tend not to be interested in C#.
nitwit005··on Ask HN: What would make you trust an AI's assessment of your understanding?
You're saying "not just completing a test", but this will inevitably be a test, just in conversation form.
nitwit005··on People who can't picture anything are rewriting the science of imagination
Watch people draw, and you'll realize that can't be true, or at least is a very incomplete description of what's going on.

People have processes they go through to eventually get to something reasonably correct. Systems for drawing people, perspective rules, etc. These get extremely elaborate. They shouldn't need any of it if they could fully visualize the scene. They'd just need to copy that image to paper.

You'll hear people say things like "trust the process". They are struggling with what they're drawing looking wrong to them, until they're fairly far along and the shapes are fairly complete. Even when they're doing things correctly, their brain isn't recognizing it.

nitwit005··on People who can't picture anything are rewriting the science of imagination
I never find the ability to visualize things, or imagine voices or sound, too interesting, as I have never used those abilities to do anything practical. I do not rehearse words in my head before talking, or create an image in my head before drawing art.

Doing art teaches you that you only have a general sense of what a lot of objects look like. People need reference for almost everything they haven't drawn before, and can struggle to understand the proportions of things they're directly looking at.

nitwit005··on Patients Who Fight Health Insurance Denials Often Win
Most of the listed problems are issues of incentives, and those are true everywhere. Insurance companies are always going to be reluctant to pay.

That there are no billboard ads for personal injury lawyers, clearly doesn't imply injury lawsuits aren't occurring.

nitwit005··on I spent $220 on Google app ads and 60% of the installs were robots
The general problem of how to advertise a small business has never been solved.

I see ads for some local pizza places, as they've learned that can work for a delivery business, but most restaurants have no ads. They don't know of anything that works.

nitwit005··on Kenyans Did College Students' Homework for Years. Then A.I. Arrived
Periodically there's some event that reveals how many people are cheating in some class, and it's never everyone.
nitwit005··on Muse – Meta’s personal AI agent
This feels poorly marketed. The examples like health, travel, and kid's education all have a ton of anxiety around them. Highlighting a confirm button existing is recognizing the problem, but isn't a fix.

You can see this not working in existing products. Alexa could book a flight for you in 2018. I've never heard of anyone using that feature. Travel planning can be extremely complex, so it's challenging to make people trust a tool is going to do the right thing.

Also, I know this is picky, but your mascot should not feel like it'd feature in a list of worst Olympic mascots.

nitwit005··on Recreating Minecraft Is Not a Benchmark
Maybe, but everyone trying to do real work is going to want to match a specific artstyle.
nitwit005··on The death of San Francisco's Market Street
A bit south of SF, the city of San Mateo has kept downtown B street, their main downtown road, closed to all traffic (https://cityofsanmateo.org/4448/B-St-Pedestrian-Mall-Improve...). That area seems just as popular as before. You still park in the same garages.
nitwit005··on ChatGPT ad targeting is garbage
I might recognize a color theme or similar. I had a friend make a joke about a lawyer who was apparently on multiple billboards for years. The name didn't click at all, so he had to explain. I did notice later due to the joke priming me.
nitwit005··on ChatGPT ad targeting is garbage
If you asked me to name the companies on the billboards that line the highway I commute on, I couldn't do it.

I don't think any single form of advertising is effective on the whole population.

nitwit005··on Uber is laying off 10% of staff
It's logical if you view them as unsuccessful.

But if "decision-making rights are unclear" is listed, it seems unlikely those people were the problem.

nitwit005··on How accurate have Ed Zitron's AI skeptic predictions been?
The most successful use of AI has almost certainly been advertising, which companies were already doing before the recent AI hype.

If by AI we mean LLMs, the question looks a bit different.

nitwit005··on When Russia Colonised California
People used to recognize the Spanish influence on California. It influenced architecture, and people understood the context of things like Zorro (https://en.wikipedia.org/wiki/Zorro). But that's now basically gone, beyond students being sent on school trips historic Spanish missions.
nitwit005··on Java is memory efficient [audio]
I agree with all the general principles, but the details of Java do matter here. There is a reason they're working hard on Project Valhalla: https://openjdk.org/projects/valhalla/
nitwit005··on Dwarf Fortress' creator says the industry's in shambles over AI
> The problem for the CEO is that every other CEO will be able to do the same thing, so what will be the distinguishing factor?

No, the problem is, no one will need the CEO. They'll press the button on their own.

nitwit005··on Dwarf Fortress' creator says the industry's in shambles over AI
Excepting strategy games, the industry's experience has been that most players won't notice clever AI.

There have been a number of games that have tried AI conversations, but you're basically trying to sell a Chat GPT prompt, and many players are aware they don't have to pay you to get a similar result on their own.

nitwit005··on My experience has nuance, yours is a data point
> Yet we turn around in our professional contexts and see numbers like “142,432 users gave this a ‘thumbs up’” and we forget everything we knew about our own individual experience with these systems.

The data people know this, the executives do not. Or rather, they choose to ignore it, because they prefer metrics be as simple as possible.

They even try to reduce numeric values to boolean ones. When you get asked on a survey to rate from "Very Unsatisfied" to "Very Satisfied", they often reduce that to a percentage count of people who answered the top 1 or 2 boxes.

I worked at a medical company where the CEO was very enthused with the idea of having large thumbs up and down buttons at the front desk, right in front of the staff. What worked for convincing him this was a bad idea was a story about staff at a retail store panicking because a child started banging on the thumbs down button. Customers got an email with thumbs up and down instead.

nitwit005··on Ask HN: What would happen if your company stopped using all AI tomorrow?
I'm fairly confident our management has managed to spend more on failed or pointless AI efforts than AI could produce in benefits.

If you tell every manager they need to show some sort of success with AI, you get a lot of projects that don't make sense, which then quietly die, and the process repeats.

nitwit005··on A society that runs on stimulants is badly broken
Even animals will do things like get drunk on fermented fruit. Humans just figured out how to get a regular supply.
nitwit005··on CEO fired developers to make room for AI. Developers create open source AI CEO
> Setting a strong, non-consensus vision. -> AI's are trained to produce the median / mode answer. So this almost disqualifies them by default.

Most human CEOs are disqualified then. Whenever I join a company these days, they stress how important the company's values are, and they are somehow basically the same as every other company.

nitwit005··on Why some US restaurants are banning tips
As with many other products, they do have an incentive to claim you're receiving something special.

Mostly yes, there's not much extra they can deliver besides a better staff to customer ratio.

nitwit005··on Where did all the public bathrooms go?
Several bathrooms at the openings to hiking trails near me had been smeared with feces. The local newspaper reported someone was caught and taken to a homeless shelter. I remain suspicious that simply resulted in the shelter's bathroom meeting the same fate.
nitwit005··on Anthropic's best AI model struggles to attract users as cheaper tools thrive
Between human languages? The problem will be Google is giving that away for free.
← PreviousPage 2 of 34Next →