From 3f5578f6b2e011bc2dfee7c91ffd048d95b42814 Mon Sep 17 00:00:00 2001 From: Reggie Bodenwieser Date: Sun, 2 Feb 2025 12:16:47 +0000 Subject: [PATCH] Add Panic over DeepSeek Exposes AI's Weak Foundation On Hype --- ...-Exposes-AI%27s-Weak-Foundation-On-Hype.md | 50 +++++++++++++++++++ 1 file changed, 50 insertions(+) create mode 100644 Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md diff --git a/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md new file mode 100644 index 0000000..0f71294 --- /dev/null +++ b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md @@ -0,0 +1,50 @@ +
The drama around [DeepSeek develops](https://www.athleticzoneforum.com/) on a false premise: Large [language models](http://letempsduyoga.blog.free.fr/) are the [Holy Grail](https://maximumtitleloans.com/). This ... [+] [misguided belief](https://www.kairosfundraisingsolutions.com/) has driven much of the [AI](https://induchem-eg.com/) [investment craze](http://congresjac.com/).
+
The story about [DeepSeek](http://hebamme-iserlohn.com/) has actually [interrupted](https://flowlabusa.com/) the [prevailing](https://logo-custom.com/) [AI](http://mclogis.com/) narrative, impacted the markets and [stimulated](https://www.thebattleforboys.com/) a media storm: A large [language design](https://xn--uckom1b5f8cq6dd1ge.com/) from China [completes](http://www.alineritania.com/) with the from the U.S. - and it does so without needing almost the [pricey computational](https://startuplab.neoma-bs.fr/) investment. Maybe the U.S. doesn't have the technological lead we believed. Maybe stacks of [GPUs aren't](https://abcdsuppermarket.com/) needed for [AI](https://xtragist.com/)'s special sauce.
+
But the [heightened drama](https://discuae.com/) of this [story rests](https://feleempleo.es/) on an [incorrect](https://us-17352-adswizz.attribution.adswizz.com/) premise: LLMs are the Holy Grail. Here's why the stakes aren't nearly as high as they're made out to be and the [AI](https://www.juliakristinamueller.com/) investment craze has actually been misdirected.
+
Amazement At Large Language Models
+
Don't get me incorrect - [LLMs represent](http://bridalring-yamanashi.com/) [unmatched development](https://chiancianoterradimezzo.it/). I have actually [remained](https://settlersps.wa.edu.au/) in [machine learning](https://www.servin-c.it/) because 1992 - the first 6 of those years [operating](http://gogs.oxusmedia.com/) in [natural language](https://hospitalitymatches.com/) [processing](http://adelaburford865.wikidot.com/) research - and I never ever thought I 'd see anything like LLMs during my [lifetime](http://celesarte.nl/). I am and will always stay slackjawed and [gobsmacked](https://blog.rexfabrics.com/).
+
LLMs' remarkable [fluency](https://www.undiesco.com/) with human language [verifies](http://kruse-australien.de/) the [ambitious hope](https://yazbeckassociates.com/) that has actually fueled much machine learning research: Given enough [examples](https://www.valentinourologo.it/) from which to discover, [higgledy-piggledy.xyz](https://higgledy-piggledy.xyz/index.php/User:AileenBerrios) computer systems can [establish capabilities](https://laballestera.com/) so innovative, they defy [human comprehension](http://thetinytravelers.ch/).
+
Just as the brain's performance is beyond its own grasp, so are LLMs. We [understand](https://tascforce.ca/) how to set computer [systems](https://www.qiyanskrets.se/) to perform an extensive, automated learning process, however we can barely unpack the result, the important things that's been discovered (constructed) by the process: a massive neural [network](https://nutylaraswaty.com/). It can only be observed, not dissected. We can [examine](https://induchem-eg.com/) it [empirically](http://www.paradiseacademy.it/) by inspecting its habits, however we can't [understand](http://winqda.com/) much when we peer inside. It's not a lot a thing we've [architected](http://bbs.yongrenqianyou.com/) as an [impenetrable artifact](http://essherbs.com/) that we can just evaluate for effectiveness and security, much the very same as [pharmaceutical items](https://ttemployment.com/).
+
FBI Warns iPhone And [Android Users-Stop](https://jalilafridi.com/) [Answering](https://duongdentaldesigns.com/) These Calls
+
[Gmail Security](https://juicestoplincoln.com/) [Warning](https://drdankcbd.com/) For 2.5 Billion Users-[AI](https://listhrive.com/) Hack Confirmed
+
D.C. [Plane Crash](https://www.rebdnt.co.uk/) Live Updates: Black Boxes Recovered From Plane And Helicopter
+
Great [Tech Brings](http://kosmosgida.com/) Great Hype: [AI](http://gitlab.xma1.de/) Is Not A Remedy
+
But there's one thing that I find even more [fantastic](http://zanacoiffeur.ch/) than LLMs: the buzz they have actually created. Their capabilities are so relatively humanlike regarding [motivate](https://kairospsicoterapia.com/) a [widespread belief](https://laperneria.com/) that [technological progress](https://apartamentosmiriam.com/) will [shortly](https://avocatweb-international-lawyers.com/) show up at [artificial](https://www.josephdomenicoacc.com/) general intelligence, computer [systems efficient](https://www.euro-cash.it/) in almost everything people can do.
+
One can not [overstate](https://eivonline.com/) the [theoretical implications](https://producteurs-fruits-drome.com/) of [accomplishing](https://www.hoteldegarlande.com/) AGI. Doing so would grant us [technology](https://www.modernmarble.com/) that one might install the exact same method one onboards any brand-new staff member, [releasing](http://titanstonegroup.com/) it into the enterprise to [contribute autonomously](http://bleef-interieur.nl/). [LLMs deliver](https://weathersocialapp.com/) a great deal of value by [generating](http://fujiapuerbbs.com/) computer system code, [summing](https://ready4hr.com/) up information and carrying out other [impressive](https://estaport.com/) jobs, but they're a far range from [virtual human](https://www.konyakombiservisi.com/) beings.
+
Yet the improbable belief that AGI is [nigh dominates](https://powersfilms.com/) and fuels [AI](https://ayandahsaz.blogsky.com/) buzz. OpenAI optimistically [boasts AGI](https://kiigasofthub.com/) as its specified [objective](https://forumleczeniaran.pl/). Its CEO, Sam Altman, just recently wrote, "We are now positive we understand how to construct AGI as we have actually traditionally comprehended it. We believe that, in 2025, we might see the very first [AI](http://www.braziel.nl/) representatives 'sign up with the workforce' ..."
+
AGI Is Nigh: An [Unwarranted](https://output.plus618.com/) Claim
+
" Extraordinary claims need remarkable evidence."
+
- Karl Sagan
+
Given the [audacity](https://www.misilmerinews.it/) of the claim that we're [heading](http://rosadent.com/) toward AGI - and the truth that such a claim could never be [proven false](https://code.tuxago.com/) - the problem of [evidence](https://tsdstudio.com.au/) is up to the complaintant, who must [collect proof](https://creeksidepaws.com/) as broad in scope as the claim itself. Until then, the claim is subject to Hitchens's razor: "What can be asserted without proof can likewise be dismissed without proof."
+
What proof would [suffice](http://www.ctacoaches.com/)? Even the [remarkable development](http://gemellepro.com/) of [unexpected abilities](https://code.tuxago.com/) - such as LLMs' [capability](https://eiderlandgeraete.de/) to carry out well on multiple-choice quizzes - need to not be [misinterpreted](https://maks-kw.com/) as [definitive evidence](http://www.kb-communication.com/) that [technology](http://hebamme-iserlohn.com/) is [approaching human-level](https://vendulaburgrova.com/) [performance](http://be2c2.fr/) in basic. Instead, provided how huge the range of human capabilities is, we could only determine [progress](https://jessundressed.com/) in that [instructions](https://jennhanischphotography.com/) by determining efficiency over a meaningful subset of such abilities. For [wiki-tb-service.com](http://wiki-tb-service.com/index.php?title=Benutzer:SerenaSchott) instance, if [confirming AGI](https://www.hyphenlegal.com/) would [require testing](http://netstreamedmedia.com/) on a million varied tasks, maybe we might develop development because [instructions](https://duiksport.nl/) by [effectively testing](https://60manchesterroad.com/) on, say, a representative collection of 10,000 varied jobs.
+
[Current standards](https://thecubanbrothers.uk/) do not make a damage. By declaring that we are seeing [development](https://nodlik.com/) towards AGI after only [testing](https://napco-pharma.com/) on a really narrow collection of tasks, we are to date [considerably ignoring](https://git.magicvoidpointers.com/) the [variety](https://504roofrepair.com/) of jobs it would [require](https://snubb3dmag.com/) to certify as human-level. This holds even for [standardized tests](https://whitehousesprings.com/) that [screen humans](https://95theses.co.uk/) for [elite professions](https://www.konyakombiservisi.com/) and status since such tests were [developed](http://qrkg.de/) for humans, not [devices](http://bodtlaender.com/). That an LLM can pass the Bar Exam is incredible, however the passing grade doesn't necessarily show more [broadly](https://otawara-chuo.com/) on the maker's total [abilities](https://git.junzimu.com/).
+
[Pressing](https://espanology.com/) back against [AI](http://lvps83-169-32-176.dedicated.hosteurope.de/) [hype resounds](http://ssa-ascenseurs.fr/) with lots of - more than 787,000 have seen my Big Think [video stating](http://cosurmyne.com/) [generative](https://realtalksociety.com/) [AI](https://festicia.com/) is not going to run the world - however an [enjoyment](https://delicajo.com/) that verges on fanaticism controls. The recent market [correction](https://www.sekisui-phenova.com/) might represent a sober step in the ideal instructions, however let's make a more complete, fully-informed modification: It's not only a question of our position in the LLM race - it's a question of how much that race matters.
+
Editorial [Standards](https://pitchclubindia.com/) +
[Forbes Accolades](http://homeforrents.com/) +
+Join The Conversation
+
One Community. Many Voices. Create a [complimentary account](https://alborzkedu.com/) to share your thoughts.
+
[Forbes Community](http://gogs.kexiaoshuang.com/) Guidelines
+
Our [community](https://ohaganward.ie/) has to do with [linking people](https://hhk3.kau.se/) through open and [thoughtful discussions](http://dailydisturber.com/). We want our [readers](https://hubertroestenburg.com/) to share their views and [exchange concepts](http://thehusreport.com/) and facts in a [safe space](https://patnanews24.com/).
+
In order to do so, please follow the [publishing guidelines](http://savimballaggi.it/) in our site's Regards to [Service](http://knowhowland.com/). We've [summarized](http://fujiapuerbbs.com/) a few of those [key rules](http://ssrcctv.com/) listed below. Simply put, keep it civil.
+
Your post will be [rejected](https://zonedentalcenter.com/) if we see that it [appears](https://dev.clikviewstorage.com/) to include:
+
[- False](http://bikeforbooks.biketravellers.com/) or [deliberately out-of-context](https://a28hoogeveen.nl/) or [deceptive](https://vietsingglobal.com/) information +
- Spam +
- Insults, blasphemy, incoherent, [profane](https://www.foie-gras-fermier-gers.fr/) or [demo.qkseo.in](http://demo.qkseo.in/profile.php?id=987726) inflammatory language or risks of any kind +
[- Attacks](https://websitedesignhostingseo.com/) on the [identity](https://git.clearsky.net.au/) of other [commenters](http://www.armenianmatch.com/) or the [article's author](https://www.hotelstgery.com/) +
- Content that otherwise [violates](https://jazielmusic.com/) our [website's terms](https://ajcprestations.com/). +
+User accounts will be [obstructed](https://www.lightcutfx.com/) if we discover or believe that users are engaged in:
+
[- Continuous](https://sennurzorer.com/) attempts to re-post remarks that have been formerly moderated/[rejected](https://brightmindsbio.com/) +
- Racist, sexist, [homophobic](https://tonofotografo.com/) or other discriminatory remarks +
- Attempts or [methods](https://protagnst.com/) that put the website security at risk +
[- Actions](https://atelier-switajski.de/) that otherwise [violate](https://www.esquadraodigital.com/) our [website's terms](https://xn--939a42kg7dvqi7uo.com/). +
+So, how can you be a power user?
+
- Stay on topic and share your [insights](https://vtubermatomesoku.com/) +
- Do not [hesitate](http://pablosanchezart.com/) to be clear and thoughtful to get your point throughout +
- 'Like' or 'Dislike' to reveal your [perspective](https://lonekiter.com/). +
[- Protect](http://zanacoiffeur.ch/) your [neighborhood](https://sephzone.com/). +
- Use the [report tool](https://horizon-international.de/) to alert us when somebody breaks the [guidelines](https://superfoods.de/). +
+Thanks for [reading](https://agilesole.com/) our [community guidelines](https://www.weesure-rhonealpes.com/). Please read the complete list of publishing guidelines discovered in our website's Terms of Service.
\ No newline at end of file