TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Ask HN: Why are so many of you so critical of AI?

15 pointsby maxdoopabout 2 years ago
Open any AI &#x2F; LLM thread on Hacker News, and within minutes you&#x27;ll get one of the following:<p>&quot;It&#x27;s a parrot.&quot;<p>&quot;It&#x27;s not actually thinking.&quot;<p>&quot;AI fans are just the new crypto-bro.&quot;<p>For such a &quot;technical&quot; and &quot;smart&quot; cohort, I find it absolutely insane that so many on HN are so overly critical; so lacking in foresight (&quot;I asked it to do math and it failed; it&#x27;ll never be helpful, ever&quot;); and so dismissive -- I truly question if it&#x27;s some form of denial.<p>What am I getting wrong? How is progress in AI as of late _not_ mind-blowing? How are we not excited at looking potential similarities between human cognition and current AI advancements?<p>I am well aware that I myself am taken with the latest advancements in the space, but I&#x27;m really trying to separate that bias here. How is it that, when Stuart Russel or Hinton come out saying, &quot;Hey, this stuff is pretty crazy and we should be careful&quot;, the vast majority of HN comments are, &quot;wow, Hinton&#x27;s Google RSUs must&#x27;ve vested; he just has nothing new to say.&quot;

19 comments

Klonoarabout 2 years ago
<i>&gt;How is progress in AI as of late _not_ mind-blowing</i><p>It <i>is</i> mind-blowing. This has absolutely no bearing on whether anyone should be critical of it or not, though.<p><i>&gt;For such a &quot;technical&quot; and &quot;smart&quot; cohort</i><p>You are attempting to imply that HN should de-facto align with AI by appealing to ego, but that doesn&#x27;t really work here - it is in fact that &quot;technical&quot;&#x2F;&quot;smart&quot; cohort that leads to the healthy criticism.<p>I would wager there <i>is</i> an issue wherein healthy criticism has started to get drowned out, and if your point was that - rather than trying to say it&#x27;s just denial - you might have an interesting discussion.
评论 #35779045 未加载
jollyllamaabout 2 years ago
Personally, I&#x27;m sick of the cheapening of everything brought about by <i>careless</i> (rather than careful) adoption of technology, and I see AI adoption as an extension of that.
评论 #35791415 未加载
sharemywinabout 2 years ago
My comments are about ChatGPT and LLMs specifically. Stable Diffusion type AIs seem like they ruin stock photography but make getting an image way cheaper so I&#x27;m not sure how to feel about that.<p>As for ChatGPT: I just don&#x27;t use it that much anymore. I played with it alot. and it was kinda fun, but when I tried to do anything much more productive with it, it just wasn&#x27;t that much help.<p>LangChain and AgentGPT looked more interesting, but when I watched some videos about them on youtube they didn&#x27;t seem to add accomplish much and got stuck in loops.<p>The biggest problem is it&#x27;s just not that reliable.<p>I think there&#x27;s an last mile type problem with it.<p>Also, there&#x27;s a governance issue with it. Who gets to decide how to use it or how it works. I&#x27;ve seen no democratic options for it. And a free un-encumbered version seems more dangerous than a corporate controlled one.
评论 #35784515 未加载
simonblackabout 2 years ago
It&#x27;s not AI that some of us are critical of. It&#x27;s the hysteria around AI that we&#x27;re critical of.<p>Yes, AI is progressing. Yes, it&#x27;s good and getting better. But &#x27;AI&#x27; is the &#x27;fad of the day&#x27; at present, just like all the previous fads-of-the-day were everywhere in the media for a while, and then just faded out of sight. &#x27;Faded out of sight&#x27; doesn&#x27;t mean they went away, they merely took their rightful place in the hierarchy of science and were no longer so much &#x27;in your face&#x27;. Hint: When did you last hear about one of history&#x27;s previous fads, &#x27;Fuzzy Logic&#x27;? It&#x27;s still around, but not everywhere in the media.
评论 #35786836 未加载
JoeAltmaierabout 2 years ago
I find AI responses to be fatuous, shallow and uncritical. It is uncreative (writes doggerel verse; doddles out prose that sounds like the requested style but only superficially resembles it), it chooses middle ground answers even when asked to do more.<p>If you need a 6th-grader to write some copy for you, then go ahead. But I&#x27;d choose an actual 6th-grader instead. They could use the practice, and they will get better with time.
评论 #35779432 未加载
al2o3crabout 2 years ago
Better question: why are YOU so invested in everybody else being excited?<p><pre><code> so dismissive -- I truly question if it&#x27;s some form of denial. </code></pre> If you don&#x27;t want people to think of you like a cryptobro, not talking like one would be a good start.
评论 #35779443 未加载
nabla9about 2 years ago
You are asking about sentiment, not about substance or anything specific.<p>&gt; &quot;It&#x27;s a parrot.&quot;<p>Analyzing the argument why or why it&#x27;s just a parrot is important. If you are not interested in the argument itself, just upset about the conclusion you are just engaging some kind of techno hype mini-religion.<p>Turning discussions into general sentiments or mental states is not beneficial unless you are in marketing.
quirkotabout 2 years ago
&gt;How is progress in AI as of late _not_ mind-blowing?<p>Yes. This is an incredible new development much in the way the atomic bomb was. Except everyone saw the atomic bomb as very, very dangerous and should be handled very very carefully. For LLMs they&#x27;re just sprouting up on every website like it&#x27;s no big deal
smoldesuabout 2 years ago
It&#x27;s neat.<p>Neatness is not always gamechanging, though. We&#x27;ve had glasses-free 3D displays available at low-prices for decades, but... it&#x27;s not a great interaction paradigm. It&#x27;s neat, to be sure - but sucks for usability.<p>That&#x27;s where I&#x27;m at with LLMs right now. I&#x27;ve watched this field evolve since Talk-To-Transformer came out, and while AI is &quot;better&quot; it has still failed to address it&#x27;s serious pitfalls (cannot cite sources, is often wrong, will confidently take extremist positions, etc.). It&#x27;s genuinely feeling like a re-hash of the cryptocurrency situation where the whitepaper is nerd porn, but the implementation is a horror story. Judging from the overall negative public sentiment of AI (even from fearmongering) I&#x27;d wager most people aren&#x27;t really interested.<p>When you zoom out and look at it from a business perspective, paying $0.02 to get your fortune told by ChatGPT is not worth it. It&#x27;s barely more useful with a less-accurate, dumber local version that&#x27;s a 4gb download. Might pan out to something interesting in the future, but so-far both language models and image generation models have disappointed me. They also haven&#x27;t really changed society that much, at least from where I&#x27;m standing.
评论 #35779492 未加载
jqpabc123about 2 years ago
<i>How is progress in AI as of late _not_ mind-blowing?</i><p>The ability to generate plausible language is impressive.<p>The fact that so much of the generated language is inaccurate and unreliable is significantly less so.<p>Basically, we have automated the art of building a talking database ... along with the art of bullshit and mimicry.
ActorNightlyabout 2 years ago
&gt;How is progress in AI as of late _not_ mind-blowing?<p>Depends on what your background is. Replace &quot;AI&quot; with &quot;information compression algorithm with semantic language lookup capability&quot;, and the endeavor still sounds cool, but not as sexy as AI.<p>Also, the current &quot;control&quot; of how these LLMs are being trained and deployed is not really that exciting, its even disheartening on some levels tbh.<p>What needs to happen is that AMD and Nvidia (and possibly others) should be racing to the bottom to create the best bang&#x2F;buck ML accelerators, where people can realistically buy racks to run in their house and transfer train these architectures on their own data sets.
anenefanabout 2 years ago
AI has the potential to be the next greatest tool. I can be certain nearly every <i>cookie cutter</i> type job in years to come, maybe a couple of decades, if not sooner, will see AI take over most of those duties, and in support, possibly do a better job if all the replies are already known.<p>I don&#x27;t personally use it or have a need at the moment, but await the day it&#x27;s matured enough I can set it on an old scanned pdf, and it&#x27;ll work out how to code a new pdf with the same type font, either by discovery or create a new font, typeset the same position, if of duplicator (ditto machine) nature, work out any miss formed letters ... also clean up any images within, such it is as good as if not better than what was originally scanned if it were printed, and greatly reduce my disc space required to store my collection of pdfs.<p>Sadly I already know that won&#x27;t become the norm for the average user, unless for some weird instance, the AI is totally under their own control - and that situation would be doubtful due to potential misuse.<p>That&#x27;s would be an example of well placed concern ... unseen misuse slowly addressed.<p>Then there will be the monetisation and I imagine it&#x27;ll creep in just like search engines, start, get better, become great and then ... fuck everyone over so those that really need a good search, pay to try others. (Sadly for any potential new web crawlers, in that time span from the end of 90s where it was open to all, there&#x27;s no simple crawling all the web any more ... )<p>The next I see as an issue will be early adopters trying to wrangle a customer base around some inappropriate use of it. New tech is nice and some people are scared of things which haven&#x27;t been well tested, but sometimes that fear is well founded - I only have to think back to voice identification technology which was really interesting in those times, but I know of at least one govt organisation that inflicted it on a wide user base despite being told it wasn&#x27;t that secure - assuming most would be hackers would not have large enough systems to mimic a voice on the fly, isn&#x27;t a good start IMO.
bckygldstnabout 2 years ago
I opened Google Maps at dinner time last night and searched &quot;Chinese&quot;, it took me to the country of China.<p>Also yesterday I asked Siri &quot;what&#x27;s 1300 mils in ounces&quot;. The text-to-speech worked properly, but instead of giving me the answer it opened a web search for &quot;1300ml in oz&quot;.<p>These are the same companies in the same market as those gearing up to dive my car and interpret my blood tests.<p>We&#x27;ve had the technology for a decade to correctly process my AI queries from yesterday. But there&#x27;s a chasm between technology and helping users, and crossing that gap is rarely the way to maximise shareholder value.
nashashmiabout 2 years ago
It has been given more credit than it deserves.<p>I understand it to be some thing that understands communication and can answer questions based on the communication it has.<p>There are some questionable capabilities like an appearance of logic. But I am going to uphold a belief that it is borrowing some standard information that it already has and manipulating it to fit the question.<p>There are areas that are logically assessable, but I do not think an LLM alone can do that. The next frontier of AI will be trying to prioritize which piece of logic is more applicable, assuming all pieces of logic are already invented.
codegeekabout 2 years ago
Because of the hype it is generating and notice how everyone and their mother is working on a product&#x2F;service around &quot;AI&quot; trying to make quick cash. It is a good technology but when you overhype something that much, it brings the worst type of people and I can guess that majority of the Crypto Bros are now doing AI.
kazinatorabout 2 years ago
GPT4 shows itself quite capable at generating code and making complex refactorings in the face of changing requirements.<p>When you stray from areas of code, it shows serious weaknesses.<p>Within still the same domain, if you discuss algorithms, it&#x27;s a little bit out to sea without a rudder. If you say subtly wrong things in your questions, it picks up on those things and pretends it researched them.<p>Earlier today, it spewed that the Lomuto partitioning scheme used in some quicksort implementations requires one bit of external storage per array element. This is completely false, and comes from a remark I made in an earlier question in the same chat about some possible idea requiring one bit of storage.<p>Like a parrot, it just took my idea, stripped it of context, and planted it into statements about something else.<p>Speaking of which, I&#x27;ve noticed a pattern in its behavior of using my ideas and hints, and pretending that it came up with them itself. If confronted about that, it will admit, that yes, sorry it did get that answer from your hint and not from its knowledge and analysis as its answer glibly implies.<p>GPT4 is intensively being trained on coding in order to impress programmers, who will then evangelize it to non-programmers. (The things it does with code really <i>are</i> impressive.)
cratermoonabout 2 years ago
&gt; How is progress in AI as of late _not_ mind-blowing?<p>What is mind-blowing about it? How is it actually thinking, and not a parrot? How do you separate the hype from the specific facts of the actual advances?<p>In other words, what are the legitimate criticisms, and if you have none, why not?
评论 #35779164 未加载
rainytuesdayabout 2 years ago
Don&#x27;t worry about it; more alpha for you.
nicbouabout 2 years ago
Unlike crypto, I <i>know</i> that AI will change the game.<p>However, I know that changing the game, inevitable as it is, does not mean making the world better.<p>We&#x27;re already seeing AI being adopted by the people who ruin things. LLMs are bullshit machines, and bullshit people are using it to make bullshit faster, cheaper and more convincing.<p>Yet AI must be trained on original art and original research. Our work. It just does so without giving credit, or a way to opt out. My work is used to train bullshit machines that will be turned on me, to profit someone else.
评论 #35788738 未加载