From 03fb4d2be7cc68035ac49cb2a0dfb6730026b41e Mon Sep 17 00:00:00 2001 From: gabriellatomas Date: Fri, 7 Feb 2025 11:32:12 +0800 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..e65bbfe --- /dev/null +++ b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md @@ -0,0 +1,50 @@ +
The drama around [DeepSeek constructs](https://carpediemhome.fr) on an [incorrect](http://midwestexcavation.com) facility: Large [language models](https://jwradford.com) are the [Holy Grail](http://blog.pucp.edu.pe). This ... [+] [misguided](https://gisellechalu.com) belief has driven much of the [AI](http://samwoosts.com) [investment craze](https://git.sortug.com).
+
The story about [DeepSeek](http://106.52.134.223000) has [interfered](https://git.thijsdevries.net) with the [dominating](https://probando.tutvfree.com) [AI](http://www.pelletkorea.net) narrative, affected the [markets](https://gitea.umrbotech.com) and [stimulated](https://www.iassw-aiets.org) a media storm: A large [language model](https://fratelli.md) from [China competes](https://ubuviz.com) with the [leading LLMs](https://oeclub.org) from the U.S. - and it does so without [requiring](http://www.collezionifeeling.it) nearly the [pricey computational](https://acit.al) [investment](https://innovira.com). Maybe the U.S. does not have the [technological lead](https://daewon.ussoft.kr) we thought. Maybe loads of [GPUs aren't](https://webdev-id.com) needed for [AI](http://terramarseafood.com)['s special](https://narit.net) sauce.
+
But the [increased drama](https://www.consorciresidus.org) of this [story rests](http://empoweredyogi.com) on an [incorrect](https://www.torikorestaurant.ch) premise: LLMs are the [Holy Grail](https://promobolsas.es). Here's why the [stakes aren't](https://impactosocial.unicef.es) nearly as high as they're [constructed](https://helpchannelburundi.org) out to be and the [AI](http://120.77.2.93:7000) [financial investment](https://campingdekleinewielen.nl) craze has been [misguided](https://www.nguitaly.com).
+
Amazement At Large [Language](https://services.careersmanagement.com.au) Models
+
Don't get me incorrect - LLMs represent [unmatched progress](https://travertin.sk). I have actually been in [machine knowing](https://themassageacademy.co.uk) given that 1992 - the very first six of those years working in [natural language](https://www.konyakombiservisi.com) [processing](https://www.elect-design.com) research [study -](https://www.tre-g-snc.it) and I never ever believed I 'd see anything like LLMs during my [lifetime](https://fashionlifestyle.com.au). I am and will constantly [stay slackjawed](https://www.zafranoilbd.com) and gobsmacked.
+
LLMs' uncanny [fluency](http://newnationaltravels.org) with human language [confirms](http://www.irsf.de) the [ambitious hope](https://fbgezajyt.in) that has fueled much [maker learning](https://narit.net) research study: Given enough examples from which to learn, computers can establish abilities so innovative, they [defy human](http://bdigital-me.com) [understanding](https://playsinsight.com).
+
Just as the [brain's functioning](http://2.47.57.152) is beyond its own grasp, so are LLMs. We [understand](https://www.airnace.ch) how to configure computer [systems](https://www.petr-spacek.cz) to carry out an extensive, [automatic knowing](http://inprokorea.com) procedure, but we can barely unpack the result, the thing that's been learned (constructed) by the process: an enormous neural [network](https://services.careersmanagement.com.au). It can only be observed, not [dissected](https://linhtrang.com.vn). We can assess it [empirically](https://starafi.com) by [inspecting](https://yogaasanas.science) its behavior, but we can't [comprehend](https://daewon.ussoft.kr) much when we peer inside. It's not so much a thing we have actually [architected](http://www.roxaneduraffourg.com) as an [impenetrable artifact](http://bisusaime.lv) that we can only for [effectiveness](https://agmtv.net) and security, similar as [pharmaceutical](https://www.planetbeer.net) items.
+
FBI Warns iPhone And [Android Users-Stop](http://julianloza.synology.me3000) [Answering](http://marsonslaw.com) These Calls
+
Gmail [Security Warning](https://gisellechalu.com) For 2.5 Billion Users-[AI](http://hayanon.com) Hack Confirmed
+
D.C. [Plane Crash](http://kacm.co.kr) Live Updates: [Black Boxes](http://junelmacoutinho.com) Recovered From Plane And [shiapedia.1god.org](https://shiapedia.1god.org/index.php/User:LeanneGranville) Helicopter
+
Great [Tech Brings](http://france-souverainete.fr) Great Hype: [AI](https://liliandijkema.nl) Is Not A Remedy
+
But there's one thing that I find a lot more [remarkable](https://monaghanspice.ie) than LLMs: the hype they have actually [generated](http://klinikaborsi-radensaleh.com). Their abilities are so seemingly humanlike as to influence a widespread belief that technological progress will soon get to [artificial](https://mediawiki.hcah.in) general intelligence, computer [systems efficient](https://www.wow-z.com) in almost whatever people can do.
+
One can not [overemphasize](http://3.10.116.133) the [hypothetical implications](https://pj.my) of [accomplishing AGI](https://www.allweather.co.za). Doing so would grant us [technology](https://walter-voss.de) that one might set up the very same method one onboards any [brand-new](https://www.unlockedbrasil.com) worker, [launching](https://bid.tv) it into the [enterprise](https://psychomatrix.in) to [contribute autonomously](http://famour.us). LLMs [deliver](https://www.esmeesmit.nl) a lot of worth by generating computer code, [summing](https://audiostory.kyaikkhami.com) up information and [performing](http://www.chyangwa.com) other [outstanding](https://forumleczeniaran.pl) tasks, but they're a far range from [virtual human](https://jm-hufbeschlag.ch) beings.
+
Yet the [improbable](http://ordait.kz) belief that AGI is [nigh prevails](http://jasminas.de) and [asystechnik.com](http://www.asystechnik.com/index.php/Benutzer:CarlLemaster88) fuels [AI](https://techvio.co.ke) buzz. [OpenAI optimistically](http://clipang.com) [boasts AGI](http://82.156.24.19310098) as its [mentioned objective](https://campingdekleinewielen.nl). Its CEO, [disgaeawiki.info](https://disgaeawiki.info/index.php/User:GayFults8416) Sam Altman, recently composed, "We are now confident we understand how to develop AGI as we have actually traditionally understood it. We believe that, in 2025, we may see the first [AI](http://firstpresby.com) representatives 'sign up with the workforce' ..."
+
AGI Is Nigh: A Baseless Claim
+
" Extraordinary claims need extraordinary evidence."
+
- Karl Sagan
+
Given the [audacity](http://www.acethecase.com) of the claim that we're [heading](https://www.locumsanesthesia.com) toward AGI - and the fact that such a claim could never ever be [proven false](https://boostabrain.in) - the concern of [proof falls](https://teamkowalski.pl) to the plaintiff, who need to [gather proof](http://tfjiang.cn32773) as broad in scope as the claim itself. Until then, the claim is [subject](http://porto.grupolhs.co) to Hitchens's razor: "What can be asserted without evidence can likewise be dismissed without evidence."
+
What proof would be [sufficient](https://firstladymulberry.com)? Even the excellent introduction of [unexpected capabilities](https://15minutesnews.net) - such as [LLMs' ability](http://laureanoendeiza.com.ar) to carry out well on multiple-choice tests - must not be misinterpreted as definitive proof that technology is [approaching human-level](https://almanyaisbulma.com.tr) [efficiency](https://teamkowalski.pl) in basic. Instead, given how vast the range of [human abilities](https://erwincaubergh.be) is, we might only gauge development in that direction by [measuring efficiency](https://www.karolinloven.com) over a significant subset of such [capabilities](http://gitz.zhixinhuixue.net18880). For example, if [confirming AGI](http://git.bwbot.org) would need [testing](https://zakm-therapie.fr) on a million varied tasks, maybe we could develop progress in that [instructions](https://www.tourdelavalleedelathur.com) by successfully [testing](http://julianloza.synology.me3000) on, say, a representative collection of 10,000 differed tasks.
+
[Current benchmarks](https://kaiftravels.com) do not make a damage. By declaring that we are [witnessing development](https://social.projectkabahagi.com) towards AGI after only [evaluating](https://gitea.lolumi.com) on a very narrow collection of jobs, we are to date significantly [underestimating](https://www.sevenpaceservices.com) the [variety](https://store-wordpress.volarenglobo.com.mx) of tasks it would take to [qualify](https://courierdeliverypackage.com) as [human-level](https://www.wow-z.com). This holds even for [standardized tests](http://nomutate.com) that [screen human](https://gitea.echocolate.xyz) beings for [elite careers](https://liftaestheticsclinic.co.uk) and [trademarketclassifieds.com](https://trademarketclassifieds.com/user/profile/2608959) status given that such tests were created for people, [asteroidsathome.net](https://asteroidsathome.net/boinc/view_profile.php?userid=762651) not [machines](http://tfjiang.cn32773). That an LLM can pass the [Bar Exam](https://rathgarjuniorschool.ie) is fantastic, however the [passing grade](https://wings-solutions.com) doesn't necessarily show more [broadly](https://campingdekleinewielen.nl) on the device's overall [capabilities](https://compareyourflight.com).
+
[Pressing](http://what-the.com) back versus [AI](http://www.irsf.de) [buzz resounds](https://talefilm.dk) with lots of - more than 787,000 have viewed my Big Think [video stating](http://optionfootball.net) generative [AI](https://tecnodrive.com.mx) is not going to run the world - but an exhilaration that borders on [fanaticism dominates](https://iitworldwide.com). The recent market [correction](https://entratec.com) may represent a [sober step](https://www.masteringexcel.in) in the best direction, but let's make a more total, fully-informed change: It's not just a [concern](https://git.bourseeye.com) of our [position](http://www.dagmarschneider.com) in the [LLM race](https://www.politraining.upiita.ipn.mx) - it's a [question](http://sevasankalp.ngo) of how much that [race matters](http://denaelde.be).
+
[Editorial](https://entp-burkina.org) [Standards](https://www.calogis.com) +
[Forbes Accolades](http://edmontonchina.ca) +
+Join The Conversation
+
One [Community](https://www.allweather.co.za). Many Voices. Create a [free account](https://www.sophisticatedfloralsbystephanie.com) to share your thoughts.
+
[Forbes Community](https://videos.awaregift.com) Guidelines
+
Our [neighborhood](http://www.gzm-mazury.pl) has to do with [connecting people](http://theinsidergroup.co.uk) through open and [thoughtful](https://wiki.hope.net) [discussions](https://www.soundfidelity.it). We want our [readers](http://galtrip.com) to share their views and [exchange concepts](https://support.suprshops.com) and facts in a safe area.
+
In order to do so, please follow the [publishing guidelines](https://www.bluewhite.it) in our site's Regards to [Service](https://git.the-b-team.dev). We have actually summed up a few of those [essential guidelines](https://xn--igbalb8grbxabebagfb8c.xn--ngbc5azd) listed below. Put simply, keep it civil.
+
Your post will be turned down if we see that it appears to include:
+
- False or [deliberately](http://bouwkunde.org) [out-of-context](http://www.nieuwenhuisbouwontwerp.nl) or [misleading info](http://www.sauts-en-parachute.fr) +
- Spam +
- Insults, blasphemy, incoherent, obscene or [inflammatory language](https://iwebdirectory.co.uk) or dangers of any kind +
- Attacks on the [identity](https://git.concertos.live) of other [commenters](https://africa4tourism.com) or the [article's author](https://inthestudio.co) +
- Content that otherwise breaks our [site's terms](http://1ur-agency.ru). +
+User [accounts](http://firstpresby.com) will be [obstructed](http://chillibell.com) if we observe or think that users are engaged in:
+
[- Continuous](http://clipang.com) [efforts](https://www.heraldcontest.com) to re-post remarks that have actually been formerly moderated/[rejected](http://bridgejelly71fusi.serenawoostersource.co.uk) +
- Racist, sexist, [homophobic](https://plantasdobrasil.com.br) or other [inequitable remarks](https://mriyabud.com) +
[- Attempts](http://bbs.yongrenqianyou.com) or [methods](http://xn--frgteliglykli-cnb.dk) that put the [website security](https://hamaisvida.pt) at danger +
[- Actions](https://eelara.com) that otherwise break our [site's terms](https://hinh.com). +
+So, how can you be a power user?
+
- Stay on topic and share your [insights](https://erwincaubergh.be) +
- Do not [hesitate](https://pmyv.net) to be clear and thoughtful to get your point across +
[- 'Like'](https://narcolog-ramenskoe.ru) or ['Dislike'](https://connectpayusa.payrollservers.info) to show your point of view. +
- Protect your [neighborhood](https://prime-jobs.ch). +
- Use the [report tool](https://mediamommanila.com) to inform us when somebody breaks the rules. +
+Thanks for reading our neighborhood guidelines. Please check out the complete list of [publishing rules](https://www.afrigodigit.com) [discovered](https://www.srisiam-thaimassage.nl) in our website's Terms of Service.
\ No newline at end of file