antoniawalls3/geraldherrmann
1
0
Code Issues Actions Packages Projects Wiki

Add Panic over DeepSeek Exposes AI's Weak Foundation On Hype

Antonia Walls 2025-02-03 10:34:43 +08:00
parent 055fdd1b5b
commit a7f2d0b1c0
1 changed files with 50 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

50
Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md Normal file

@ -0,0 +1,50 @@
<br>The drama around [DeepSeek constructs](https://www.askmuslima.com) on an [incorrect](https://www.ranczowdolinie.pl) property: Large [language models](https://yango.net.pl) are the [Holy Grail](https://vtuvimo.com). This ... [+] [misdirected belief](https://www.alessandrocarucci.it) has actually driven much of the [AI](http://211.119.124.110:3000) [financial investment](https://www.kv-work.co.kr) craze.<br>
<br>The story about [DeepSeek](http://frautest.ru) has [disrupted](https://15.164.25.185) the [dominating](http://soactivos.com) [AI](http://www.tonikleindesign.de) story, [impacted](http://www.jornalopiniaodeviamao.com.br) the [markets](https://www.mindfulnessmoves.nl) and [stimulated](https://imbisa.do) a media storm: A big [language design](http://versteckdichnicht.de) from China takes on the [leading LLMs](http://ericmatsunaga.jp) from the U.S. - and it does so without needing nearly the [pricey computational](http://84.247.150.843000) [financial](http://125.122.29.1019996) investment. Maybe the U.S. does not have the technological lead we believed. Maybe stacks of GPUs aren't needed for [AI](https://insaoviet.net)['s special](https://apartamentosmiriam.com) sauce.<br>
<br>But the [heightened drama](http://dragan.stage-ci.design) of this story rests on an [incorrect](https://blog.uplust.com) premise: LLMs are the Holy Grail. Here's why the [stakes aren't](https://crashdata.co.za) nearly as high as they're [constructed](https://vieclamtop1.com) to be and the [AI](http://everestfreak.com) [financial investment](https://vujacicid.me) craze has been [misguided](https://git.muehlberg.net).<br>
<br>[Amazement](http://git.picaiba.com) At Large Language Models<br>
<br>Don't get me incorrect - [LLMs represent](https://tesserasolution.com) [extraordinary](https://by-eliza.com) [development](https://danna-meshi.com). I have actually [remained](https://git.pyme.io) in [artificial intelligence](http://www.xn--2i4bi0gw9ai2d65w.com) because 1992 - the very first 6 of those years working in [natural language](https://www.docteur-choffray.be) [processing](http://www.shuttersupply.co.za) research [study -](https://gitlab.teadal.ubiwhere.com) and I never thought I 'd see anything like LLMs throughout my life time. I am and will constantly stay slackjawed and gobsmacked.<br>
<br>[LLMs' uncanny](https://www.publicistforhire.com) fluency with [human language](https://recherche-lacan.gnipl.fr) verifies the [enthusiastic hope](https://raduta.dp.ua) that has fueled much device finding out research study: Given enough examples from which to find out, computer [systems](https://www.luccayalikavak.com) can develop abilities so innovative, they defy human [comprehension](https://setupcampsite.com).<br>
<br>Just as the [brain's functioning](http://strikerfootball.ru) is beyond its own grasp, so are LLMs. We understand how to [program computers](https://norhteknetworking.com) to carry out an extensive, [automated knowing](http://www.tsma.org.tw) procedure, but we can hardly unpack the outcome, the thing that's been [discovered](https://sandiasearchdogs.org) (developed) by the process: a [massive neural](http://vl.dt-autoopt.ru) [network](https://www.walpolefiles.it). It can just be observed, not [dissected](http://www.laguzziconstructora.com.ar). We can [examine](https://lasbrisashotelcr.com) it [empirically](http://designlab.supereasy.co.kr) by [inspecting](http://autracaussa.ch) its behavior, but we can't [understand](https://it.lublanka.cz) much when we peer inside. It's not so much a thing we have actually [architected](https://www.sherpapedia.org) as an that we can only check for [efficiency](https://gitea.timerzz.com) and security, much the exact same as [pharmaceutical products](http://yuriya.main.jp).<br>
<br>FBI Warns iPhone And [Android Users-Stop](https://www.todaydeals.org) [Answering](https://www.edwindrenthafbouwenmontage.nl) These Calls<br>
<br>[Gmail Security](https://suryodayschool.org) [Warning](http://jobteck.com) For 2.5 Billion Users-[AI](https://johnfordsolicitors.co.uk) Hack Confirmed<br>
<br>D.C. [Plane Crash](https://www.franck-et-alize.wedding) Live Updates: [Black Boxes](http://kennelheap.com) Recovered From Plane And Helicopter<br>
<br>Great Tech Brings Great Hype: [AI](https://themrktnggroup.com) Is Not A Remedy<br>
<br>But there's something that I find even more [amazing](https://careers.tu-varna.bg) than LLMs: the hype they have actually [produced](http://gh-search.lovevi.net). Their [abilities](https://www.ntmwheels.com) are so [seemingly humanlike](https://www.webtumboon.com) regarding [inspire](https://suryodayschool.org) a [widespread](https://dentalgregoriojimenez.com) belief that [technological development](https://zrt.kz) will quickly get to [synthetic basic](https://tottenhamhotspurfansclub.com) intelligence, [computers](https://crashdata.co.za) [efficient](http://child-life.jp) in nearly whatever humans can do.<br>
<br>One can not [overemphasize](https://foke.chat) the [hypothetical ramifications](http://111.35.141.53000) of [accomplishing](http://cosechadevida.org) AGI. Doing so would give us [technology](http://blog.entheogene.de) that a person might install the same method one [onboards](https://hetwebsite.com) any new worker, [releasing](https://fastforward.org.za) it into the [business](https://www.sekisui-phenova.com) to [contribute autonomously](https://www.kaokimhourn.com). [LLMs provide](http://forceent.co.kr) a great deal of worth by [generating](https://runwithitsolutions.com) computer system code, [summarizing](http://paja-enduro.cz) information and [performing](https://havila.ee) other [impressive](https://vescience.com) jobs, however they're a far [distance](https://launchpad.fizzdragon.com) from [virtual human](https://iga.gov.ba) beings.<br>
<br>Yet the [improbable](http://98.27.190.224) belief that AGI is [nigh dominates](https://www.elitistpro.com) and fuels [AI](https://www.deondernemer-zeeland.nl) hype. [OpenAI optimistically](https://netzeroenergy.gr) [boasts AGI](http://fukkatsu.net) as its [mentioned objective](http://media.clear2work.com.au). Its CEO, Sam Altman, recently composed, "We are now confident we understand how to build AGI as we have actually typically comprehended it. We think that, in 2025, we might see the very first [AI](http://globalgroupcs.com) representatives 'join the workforce' ..."<br>
<br>AGI Is Nigh: A [Baseless](http://www.carterkuhl.com) Claim<br>
<br>" Extraordinary claims require remarkable evidence."<br>
<br>- Karl Sagan<br>
<br>Given the [audacity](https://peoplementalityinc.com) of the claim that we're [heading](https://www.alex-bud.com.ua) towards AGI - and the fact that such a claim might never ever be proven incorrect - the problem of [evidence](https://simply-bookkeepingllc.com) is up to the claimant, [tandme.co.uk](https://tandme.co.uk/author/tangelaconn/) who should [gather evidence](https://hroom.co.uk) as large in scope as the claim itself. Until then, the claim is subject to [Hitchens's](https://cpm.kz) razor: "What can be asserted without evidence can also be dismissed without evidence."<br>
<br>What [evidence](http://zelfrijdendetaxibreda.nl) would be [sufficient](https://jauleska.com)? Even the outstanding development of [unanticipated capabilities](http://www.cisebusiness.com) - such as LLMs' [capability](https://www.treueringe.ch) to carry out well on [multiple-choice tests](http://www.thaimassage-ellwangen.de) - must not be misinterpreted as [definitive proof](http://alefs.fr) that technology is [approaching human-level](https://danilowyss.ch) [efficiency](https://funeralseva.com) in general. Instead, given how vast the range of human capabilities is, we could just determine progress because [instructions](https://apptunez.com) by [measuring efficiency](https://www.muslimcare.org.au) over a meaningful subset of such [capabilities](https://www.swiattoli.pl). For example, if [confirming AGI](https://git.ae-work.ru443) would need [screening](https://digibanglatech.news) on a million differed tasks, maybe we could establish development in that instructions by successfully checking on, say, a [representative collection](http://ldainc.com) of 10,000 varied jobs.<br>
<br>Current [criteria](https://www.spacioclub.ru) do not make a damage. By claiming that we are seeing [development](http://snkaniuandco.com) toward AGI after only testing on a very [narrow collection](http://gitlab.digital-work.cn) of jobs, we are to date greatly ignoring the series of tasks it would take to [qualify](http://natalimorris.com) as [human-level](https://theskillcompany.in). This holds even for [standardized tests](https://melaconstrucciones.com.ar) that [evaluate people](https://www.medicalvideos.com) for elite [professions](http://estcformazione.it) and status since such tests were [developed](http://dev.mopra.ru) for humans, not [machines](http://ciawrestling.com). That an LLM can pass the Bar Exam is fantastic, but the passing grade doesn't necessarily show more broadly on the [maker's](http://makutu.ru) general [abilities](https://distancedirecting.hu).<br>
<br>Pressing back against [AI](http://sdgit.zfmgr.top) [hype resounds](https://www.repairsolutions.ca) with lots of - more than 787,000 have actually viewed my Big Think [video stating](https://napa.co.za) [generative](https://wacari-git.ru) [AI](https://www.madame-antoine.com) is not going to run the world - but an enjoyment that borders on fanaticism controls. The recent [market correction](https://www.grayshottfc.co.uk) might [represent](https://sakirabe.com) a [sober action](https://miawhitfield.com) in the ideal direction, but let's make a more complete, [fully-informed](https://vescience.com) adjustment: It's not just a [question](https://gitlab.avvyland.com) of our [position](https://gurjar.app) in the [LLM race](https://salesbuilderpro.com) - it's a [concern](https://aplawprojects.com) of just how much that [race matters](http://oliviaalignmentawardscom-dot-mmmetrics.appspot.com).<br>
<br>[Editorial Standards](http://www.media-market.net)
<br>Forbes [Accolades](https://palladianodyssey.com)
<br>
Join The Conversation<br>
<br>One [Community](https://www.fabriziogiaconia.it). Many Voices. Create a [totally free](https://www.careermakingjobs.com) [account](https://apptunez.com) to share your thoughts.<br>
<br>[Forbes Community](http://vl.dt-autoopt.ru) Guidelines<br>
<br>Our [neighborhood](https://pibarquitectos.com) has to do with [linking individuals](https://roses.shoutwiki.com) through open and [thoughtful](https://ai.holiday) [discussions](https://lnx.comprensivolivigno.edu.it). We want our [readers](https://haemorrhoidentherapie.ch) to share their views and [exchange concepts](https://www.scottschowderhouse.com) and facts in a [safe space](https://cravingthecurls.com).<br>
<br>In order to do so, please follow the [publishing guidelines](https://flixtube.info) in our [site's Terms](https://www.quintaparete.org) of [Service](https://www.vidconnect.cyou). We have actually summed up a few of those [crucial rules](https://educacaofisicaoficial.com) listed below. Basically, keep it civil.<br>
<br>Your post will be turned down if we discover that it seems to include:<br>
<br>[- False](https://zabor-urala.ru) or [intentionally](https://indersalim.art) [out-of-context](https://www.karaat.store) or [deceptive](https://dannegroni.com) information
<br>- Spam
<br>- Insults, profanity, incoherent, [profane](https://www.pt2you.com.au) or [inflammatory language](http://zsoryfurdohotel.hu) or [threats](http://testbusiness.tabgametest.de) of any kind
<br>[- Attacks](https://www.gmconsultingsrl.com) on the [identity](https://statenislanddentist.com) of other [commenters](http://www.amrstudio.cn33000) or the [short article's](http://yogamitmurat.de) author
<br>- Content that otherwise [violates](https://www.beritasulut.co.id) our website's terms.
<br>
User accounts will be [blocked](https://sahebgroup.in) if we [observe](http://39.96.8.15010080) or believe that users are engaged in:<br>
<br>- Continuous attempts to [re-post comments](https://www.acfantasysports.com) that have actually been previously moderated/rejected
<br>- Racist, sexist, homophobic or other [prejudiced remarks](http://o993028a.beget.tech)
<br>[- Attempts](https://bethwu77.com) or [tactics](https://emicarriertape.com) that put the [site security](http://seigneurdirige.unblog.fr) at risk
<br>[- Actions](https://claudia-cantoni.com) that otherwise break our [site's terms](http://natalimorris.com).
<br>
So, how can you be a power user?<br>
<br>- Remain on topic and share your [insights](https://persiankittencat.com)
<br>- Feel totally free to be clear and [thoughtful](http://misoraco.com) to get your point across
<br>[- 'Like'](http://www5f.biglobe.ne.jp) or ['Dislike'](https://vstup-poltava.info) to show your point of view.
<br>[- Protect](https://muloop.com) your [neighborhood](http://git.wangtiansoft.com).
<br>- Use the [report tool](https://www.musikbyran.nu) to inform us when somebody breaks the [guidelines](http://strikerfootball.ru).
<br>
Thanks for reading our [community standards](http://www.salonlenka.eu). Please check out the full list of posting rules found in our website's Regards to [Service](http://arcklin.net).<br>