Hacker Newsnew | past | comments | ask | show | jobs | submit | smallstepforman's commentslogin

Google search is giving us a taste of AI summarised results, and for simple things its passable, but ask a serious question and you get good looking garbage. Yes, I know its early days, but looking at the current output quality we have nothing to worry about. It will be used as calculators, offload some menial repetetive task which can be automated, but the next gen of developers will still be tasked to solve complex problems.

Google AI the other day told me that tinnitus is listed as a potential adverse reaction of Saphnelo.

Only it damn well isn’t. Anywhere. Not even patient reports.

The problem with AI is if it’s right 90% of the time but I have to do all the work anyway to make sure it’s not one of the 10% of times it’s extremely confidently wrong, what use is it to me?


This problem is has already gotten so much better. In my experience it's no longer 10% of the time (I'd estimate more like 1%). In the end, you still need to use judgement; maybe it doesn't matter if it's wrong, and maybe it really does. It could be citing papers, and even then you don't know if the results are reproducible.

Has it actually become that much better or have you let your standards and judgment lapse because you want to trust it?

How would you even know to evaluate that?


Ya I've had basically this question for a while. My assumption is that most of the time people search the internet to answer questions they DONT know the answer to.

If an LLM gives you a response to that question, how do you know if its right or wrong without already knowing the answer or verifying it some other way? Is everyone just assuming the ai answers are right a majority of the time? Have there been large scale verification of a wide variety of questions that I'm not aware of?


Case in point.

I purchased a small electronic device from Japan recently. The language can be changed to English, but it’s a bit of a process.

Google’s AI just copied a Reddit comment that itself was probably AI generated. It made up instructions that are completely wrong.

I had to find an actual human written web page.

The problem is with more and more AI sloop, less humans will be motivated to write. AGI at least the first generation is going to be an extremely confident entity that refuses to be wrong.

Eventually someone is going to lose a billion dollars trusting it, and it’ll set back AI by 20 years. The biggest issue with AI is it must be right.

It’s impossible for anything to always be right since it’s impossible to know everything.


Google's AI overviews is the single worst AI-driven experience in widespread use today. It's a mistake to make conclusions about how good AI has got based on that.

Have you tried search in ChatGPT with o4-mini or o3?


I don't use it that much, but I have noticed these AI overviews still seem hallucinate a lot, compared to others. Meanwhile I hear that Gemini is catching up or surpassing other models, so I wonder if I'm just unlucky (or just haven't used it enough to see how much better it is)

One is their state-of-the-art model, the other one's the best model they can run at scale and speed people expect from a search engine.

I had a couple of great examples of getting the exact opposite answer depending on how I worded my question, now I can't trust any of the answers.

Maybe the answer to your question was subjective?

The first set of questions was about code review timing in a workflow (background: was in a discussion about different styles of workflows):

Q1.1: "do most developers do code reviews before testing"

A1.1: essentially "yes most before..."

Q1.2: "do most developers do code reviews after testing"

A1.2: answer was essentially "yes most after..."

The 2nd set of questions was related to building a retaining wall. I've never used the type of wall block with a center notch and groove, only the type with a lip on the back. The center type creates a wall straight up but lip in back leans back a little. I was curious if one was more stable than the other:

Q2.1: "is retaining wall block with center groove more stable"

A2.1: essentially "yes, block with center groove is more stable..."

Q2.2: "is retaining wall block with back lip more stable..."

A2.2: essentially "yes, block with lip on back is more stable..."

When I finally found an engineering website with real details, the answer was that the lip on back was a little more stable due to the resulting angle of the wall. It also emphasized that the lip and the center groove are not factored in to stability calcs at all, they are only for alignment, gravity and friction of the blocks is what holds it in place.


It makes me wonder, why don't we want AI to be deterministic and make it nondeterministic on purpose instead? I hope someone can explain it to me...

Google Search AI is the worst and considering that AI is not a good alternative to search (the models are compressed data), I am not sure why Google has decided to use LLMs to answer questions.

Try asking Perplexity for a real taste. It works far better than google's--good enough to make searching fun again.

Try coding with Claude instead of Gemini. Those that do tell me it is well beyond.

Look at the recent US jobs reports--the draw down was mostly in professional services. According to the chief economist of ADP "Though layoffs continue to be rare, a hesitancy to hire and a reluctance to replace departing workers led to job losses last month."

Of course, correlation is not causation, but everyone white collar person I talk with is saying AI is making them far more productive. It's not a leap to figure out that management sees that as well.


Its easy to reinstall the OS. Its a lot more damaging if you lose your childs birthday photos, tax documents and anything you actually care about. This is where the entire PC security fiasco breaks down, since I want my docs directory protected FROM any system installed app/driver. I want an OS that asks for permission when accessing doc directory.

I actually had a radio with circular radio buttons, which would pop back when you selected another option. It had switches instead of check boxes.

The one that drives me crazy is slider based checkboxes. I never know which side is on/off. Bad UI convention.

And speaking of checkboxes, I want an actual tick mark (checkmark), not a X cross. Its called checkbox, not Xbox or crossbox, it has to be a checkmark. Also, its a square, not a box. Disaster.


You mean those toggles that are very common on settings pages (i.e. in Android/iOS)? If they are colored, they are very easy to parse, imo, but it never hurts to actually write "on"/"off".

Those toggles actually mimic real hardware that used to be fairly common. I find those should be preferred over checkboxes for anything that takes immediate effect. If they don't, and you're collecting a bunch of options at once, in a form, then use checkboxes.


Some toggles are labeled terribly as well, so it’s not clear what “on” even means. Or double negatives so it demands that extra mental cycle just for the sake of having all the sliders to the same side in the default configuration.

Oh yeah, those are just objectively awful X)

>If they are colored, they are very easy to parse

unless the colors are red and green, and the user happens to be red-green colorblind. So yes, always have text indicate on/off as well.


I thought about more then two options. For example when you have 10 TV or Radio Channels. They are numbered from 1 to 10. And only one channel can be chosen. Or for example, you can buy concert tickets, maximum is 4 per purchase. You may want radio buttons with a number from 1 to 4. Or you have to choose a color or size for a t-shirt (Mostly they look like buttons but there functionality is radio).

Unlabeled slider switches were never particularly common.

For instance, my old stereo has push button toggles, where “in” means “on” (this convention was common because of how those switches work), and three way levers with labels on two of the three positions (there’s no space to label the middle position, and it means “default”.


I remember them on mp3-Player, Walkmans, Microphones and even Mobile-Phones. Usually on device that you want to lock or particularly turn on and off. And sometimes you have to push them hard with the help of your Fingernail.

Often enough they are on some websites settings, with (almost) no color, but labelled with imperatives. Option X: "activate". Do I press to activate, or is it already on?

> If they are colored, they are very easy to parse

Relying on color to make something easy to parse is an awesome accessibility choice.


They can be colored and adapt to accessability settings, including color corrections for different types of colorblindness or other impairments. All the toggle designs I've seen in the wild also have the space to write "on"/"off", a check/cross, etc.

Great, now we need a 12 bit BGR game for engineers on little endian systems ;)

Every simple code base / API is faster than mature API’s due to the fact they do less. A simple string handling library which isnt biderectional, doesnt cater to people with accessibility, doesnt take locale formatting into account, doesnt cover 100% unicode spec will always be faster than complex code that does.

Code and kernels that target known hardware doesn’t need dynamic conditional code to handle unpredictable hardware. This will be faster.

General purpose operating systems handle printing events, background updates, periodic online checks, network discovery, maintenance jobs etc, all these operations consume resources and time.

Yes, Steam deck on Linux will run faster than equivalent games on Windows. But Steam deck on a smaller OS like Haiku will run even faster than Linux.

Engineering is a compromise. A F1 car can corner faster than a passanger car. But it probably sucks to reverse park. Also, I cannot imagine using a sports car for grocery shopping and hauling furniture from Ikea.


It took me 3 hours to implement A* with hex tiles, got it working on first attempt (land tiles only), specifically for Civ type game. It gets complex when you want to group units so that they travel together. Adding water crossings with cargo ships and war ships is a different challenge.


If you want cohesion between entities pathfinding, adjust the cost when you do the pathfinding for tiles that has friendlies on them to be lower than their base cost.

The way to think about water crossing with naval transports, is to consider those things to be conditions. You already have a set of condition when pathfinding. Just add another case for water. Make it so the requirement is that you’re either on a ship or there is a ship on the adjacent tile you checked previously, e.g N-1. If valid, set a flag and now every tile check that is water should be appropriate.


I considered and bought Apple hardware when they transitioned to x86_64, since I run more than OSX systems. I’m not switching to Arm since I cannot run() non OSX systems. So their personal computer line is dead to me.

() I’m aware of Asahi Linux, but that is not what I run.


They're introductory m4 is often< $500.

I'm generally not a fan but hell, that's peanuts

Got mine for $460


Computers are just getting cheap, now. You can buy an equivalently specced Ryzen Mini PC for $300.

That said, cheap Macs never really grabbed my attention and more than a cheap ChromeOS box does. I want to pay for more features, not less. I already have a media transcode server, there's really nothing I could use a Mac mini for besides maybe the world's most lame set-top box.


I got it for running AI models


I apologize in advance.


Is Windows arm also not suitable?


“ endless supply of uneducated conscripts “

Now you’re just being silly.


Regardless of how much education they have - they are treated as uneducated.


There is this wonderful presentation by Herb Sutter talking about how the C++ concept “class” covers over 20 other abstractions, and that Bjarne’s choice for C++ was the right choice since it offers so much power and flexibility and expressive power in a concise abstraction.

Other languages (just like the article) only saw the downsides to such a generic abstraction that they added N times more abstractions (so split inheritance, interfaces, traits, etc) and rules for interactions that it significantly complicated the language with fundamentally no effective gains.

In summary, Herb will always do a better job than me explaining why the choices in the design of C++ classes, even with multiple inheritance, is one of the key factors of C++ success. With cppfront, he extends this idea with metaclasses to clearly describe intent. I think he is on the right track.


Could you share a link?


Grow up, you cannot honestly think that government employees in Russian and China dont have more important things to deal with than ruining western teenage gamers experience with state funded cheat codes and hacks.


That's the thing with psyops and "plausible deniability" stuff. Do too much of it and eventually you'll be blamed for everything that's gone bad no matter if you actually did it or not. China and Russia are in the "FO" phase of "FAFO".

Besides, sowing discontent is a tried and true propaganda strategy.


I'm sorry, but no posts above make any sense, which is obvious to anyone with any degree of familiarity with chinese/russian/brazilian/german/british/north american cheating demographic and communities (yes that's a thing, also note how those two are only 1/3 of the worst offenders, and the poster above clearly doesn't have any idea about it). Stop the nationalistic flamebait and FUD, please.


Everyone cheats, I think no one here is seriously arguing that, no matter the country.

The question is: who coughs up the money for developing these cheats? Some of these, particularly the PCIe hardware rootkits, take a lot of money, time and skilled people to develop - and it is not too far fetched to assume that a nation state has been of assistance here.

Others openly flout their allegiance to Russia like the MIG-Switch developers, a ton of "bulletproof hosters" use Russian ASNs and/or are based in Russia, malware automatically disables itself when it detects indications of being in Russia... I can explain the latter away as "don't shit where you eat", but the others? There's no way there aren't direct links between the Russian government and the criminal actors. At the very least there must be some sort of "tacit approval".


I know that I'm repeating myself but none of that has any connection to reality for anyone remotely familiar with the devs of specific cheats and their history. Especially DMA hardware developers which are neither Russian nor Chinese. I can write a long post about this some day (especially about how local online game hacking economy/culture works, for every cultural bubble - they overlap a lot, and specific moments of drama inside them) and post it to HN, but please, stop being a part of the "psyop" you're talking about.


That wasn't my point though. My point was to offer an explanation why everyone is so quick to blame Russian and Chinese propaganda these days - I call it "inverse yell fire" or "inverse crying wolf": both countries have so often denied any involvement or responsibility in clear and serious violations that now everyone defaults to not believing their denials.


It could be a good infection vector. I’m under the impression that these programs are generally distributed on shady sites, need low-level access and high permissions, and are generally (actually, not sure about this part) proprietary and closed source. So, release some good cheats, and see if anybody important downloads one.

Government officials and security researchers play videogames too, after all… sure, we’d hope they’d have good computer self defense. But it is a fishing expedition and all that.


Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: