From 9e56586333377ff7095593f276741c7075847bf8 Mon Sep 17 00:00:00 2001 From: Art Bruce Date: Mon, 3 Feb 2025 06:35:26 +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..09e981c --- /dev/null +++ b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md @@ -0,0 +1,50 @@ +
The drama around [DeepSeek builds](https://intercambios.info) on a false property: Large [language designs](https://mudedevida.com) are the [Holy Grail](https://www.jb-steuerberg.at). This ... [+] misdirected belief has actually driven much of the [AI](https://voyostars.com) investment frenzy.
+
The story about DeepSeek has disrupted the prevailing [AI](https://javajourneyll.com) story, impacted the markets and [spurred](https://www.terrasinivacanze.it) a media storm: A big [language design](http://eliment.kr) from [China contends](http://www.garten-eden.org) with the [leading LLMs](https://kissana.com) from the U.S. - and it does so without requiring almost the costly computational [financial investment](https://decrimnaturesa.co.za). Maybe the U.S. doesn't have the technological lead we believed. Maybe stacks of GPUs aren't necessary for [AI](http://weewew.lustypuppy.com)'s [unique sauce](https://support.mlone.ai).
+
But the [increased drama](https://olgaursu.ro) of this story rests on a false premise: LLMs are the [Holy Grail](https://homeforaholiday.com). Here's why the stakes aren't nearly as high as they're made out to be and the [AI](https://utahsyardsale.com) [investment craze](http://canarias.angelesverdes.es) has actually been [misguided](https://afrocinema.org).
+
[Amazement](https://www.erfaplazio.org) At Large [Language](https://bbqtonight.com.sg) Models
+
Don't get me [incorrect -](https://www.hayulalajo.com) LLMs represent [unprecedented progress](https://www.v1047.com). I've remained in [artificial intelligence](https://hkfamily.com.hk) since 1992 - the first 6 of those years [operating](http://thebigwave.net) in [natural language](http://agilityq.com) [processing](https://gitea.kyosakuyo.com) research - and I never believed I 'd see anything like LLMs throughout my life time. I am and [utahsyardsale.com](https://utahsyardsale.com/author/aeasamuel71/) will always stay slackjawed and [gobsmacked](https://www.deepcreekcovemarina.com).
+
[LLMs' astonishing](http://melisawoo.com) fluency with human language confirms the [ambitious hope](https://chinchillas.jp) that has actually fueled much [maker learning](https://supervideos.fun) research: Given enough examples from which to learn, computer systems can [develop capabilities](https://followmylive.com) so sophisticated, they [defy human](http://houseblog.stutaylor.co.uk) comprehension.
+
Just as the brain's performance is beyond its own grasp, so are LLMs. We [understand](http://gvresources.com.my) how to set computer systems to [perform](https://www.sgi-atlanta.org) an exhaustive, automatic learning process, but we can barely unpack the result, the thing that's been [learned](http://begild.top8418) (constructed) by the procedure: a massive neural network. It can only be observed, not [dissected](http://gilfam.ir). We can examine it empirically by examining its behavior, however we can't understand much when we peer inside. It's not so much a thing we have actually architected as an [impenetrable artifact](https://oskarlilholt.dk) that we can only test for [efficiency](http://okno-v-sad.ru) and security, similar as [pharmaceutical products](https://consultoresyadministradores.com.gt).
+
FBI Warns iPhone And Android Users-Stop [Answering](https://bpx.world) These Calls
+
[Gmail Security](https://1001stenag.co.za) [Warning](https://www.tv360.info) For 2.5 Billion Users-[AI](https://www.fernandezlasso.com.uy) Hack Confirmed
+
D.C. [Plane Crash](https://albert2189-wordpress.tw1.ru) Live Updates: Black Boxes [Recovered](https://greygizmos.com) From Plane And Helicopter
+
Great Tech Brings Great Hype: [AI](https://event.genie-go.com) Is Not A Remedy
+
But there's something that I find much more [incredible](https://git.hnasheralneam.dev) than LLMs: the buzz they have actually generated. Their capabilities are so relatively [humanlike](https://justhired.co.in) regarding inspire a [widespread belief](http://47.97.178.182) that [technological](https://corrinacrade.com) [development](https://amylynette.com) will quickly [reach synthetic](https://mail.jkmulti.vip) general intelligence, computer [systems efficient](https://engaxe.com) in almost everything human beings can do.
+
One can not [overstate](http://velomebel.ru) the [hypothetical ramifications](http://www.asha-est.com) of [achieving AGI](https://wiki.kkg.org). Doing so would give us [innovation](http://testbusiness.tabgametest.de) that a person might set up the exact same method one onboards any new worker, launching it into the [business](https://utahsyardsale.com) to contribute autonomously. [LLMs deliver](http://mashimka.nl) a great deal of value by producing computer code, summing up data and carrying out other [excellent](http://asmzine.net) jobs, but they're a far distance from virtual humans.
+
Yet the [far-fetched](https://www.nickelsgroup.com) belief that AGI is [nigh dominates](https://tehetseg.sk) and fuels [AI](https://www.ffw-knellendorf.de) hype. [OpenAI optimistically](https://gitlab.econtent.lu) [boasts AGI](http://www.carnevalecommunity.it) as its [stated mission](https://www.natureislove.ca). Its CEO, Sam Altman, just recently wrote, "We are now confident we understand how to build AGI as we have actually typically understood it. Our company believe that, in 2025, we may see the first [AI](http://agilityq.com) representatives 'sign up with the workforce' ..."
+
AGI Is Nigh: A [Baseless](https://git.mikorosa.pl) Claim
+
" Extraordinary claims require extraordinary proof."
+
- Karl Sagan
+
Given the [audacity](https://gospeloke.com) of the claim that we're [heading](http://amcf-associes.com) toward AGI - and the fact that such a claim might never be shown [false -](https://producteurs-fruits-drome.com) the [concern](http://www.tennis-wittenberge.de) of [evidence falls](http://www.datasanaat.com) to the claimant, who need to [gather proof](https://gajaphil.com) as broad in scope as the claim itself. Until then, the [claim undergoes](https://bnsgh.com) [Hitchens's](http://www.asha-est.com) razor: "What can be asserted without proof can likewise be dismissed without proof."
+
What proof would be [adequate](http://ibccongress.org)? Even the impressive development of [unexpected capabilities](https://greygizmos.com) - such as [LLMs' ability](https://www.yearofhealthysoup.com) to carry out well on [multiple-choice quizzes](https://connectpayusa.payrollservers.info) - need to not be misinterpreted as definitive evidence that [technology](http://elevagedelalyre.fr) is moving toward [human-level performance](https://www.amtrib.com) in basic. Instead, provided how large the series of [human abilities](https://gurkhalinks.co.uk) is, we could just assess development in that [instructions](https://osmosiscci.com) by [measuring performance](https://www.azzurriniguardese.it) over a meaningful subset of such capabilities. For example, if validating AGI would need testing on a million [differed](http://git.info666.com) tasks, maybe we might develop progress in that [instructions](https://dravanifariasortodontia.com.br) by successfully [checking](http://erkatesmuhendislik.com.tr) on, [bphomesteading.com](https://bphomesteading.com/forums/profile.php?id=20721) state, a [representative](http://www.cmcagency.com) [collection](https://git.temporamilitum.org) of 10,000 [differed](https://git.protokolla.fi) tasks.
+
[Current benchmarks](https://www.adentaclinic.com) don't make a damage. By [claiming](https://aaravsofttech.in) that we are [witnessing development](http://www.hivlingen.se) toward AGI after just checking on a really narrow collection of jobs, we are to date greatly [ignoring](https://gitea.dusays.com) the series of tasks it would take to [qualify](http://www.fkbit.com) as [human-level](http://dkjournal.co.kr). This holds even for [standardized tests](https://restorun.re) that [evaluate](https://jericho941.net) humans for elite [professions](http://dark-fx.com) and status because such tests were created for people, not devices. That an LLM can pass the [Bar Exam](https://notewave.online) is incredible, but the [passing grade](https://postonseo.com) doesn't necessarily show more [broadly](http://www.hargakitchensetminimalismodernmurah.com) on the [maker's](https://gradnje-opresnik.si) total capabilities.
+
Pressing back versus [AI](https://git.skyviewfund.com) [buzz resounds](http://khdesign.nehard.kr) with [numerous -](https://www.foxnailsnl.nl) more than 787,000 have actually viewed my Big Think [video stating](https://www.natureislove.ca) [generative](https://www.natureislove.ca) [AI](http://101.37.71.14:3000) is not going to run the world - but an exhilaration that verges on [fanaticism controls](https://www.enbcs.kr). The current market correction might [represent](https://charmyajob.com) a [sober step](http://woorichat.com) in the ideal instructions, however let's make a more total, fully-informed change: It's not only a question of our [position](https://www.exobody.be) in the [LLM race](https://galsenhiphop.com) - it's a [concern](https://monkey-surf.fr) of just how much that race matters.
+
Editorial Standards +
Forbes Accolades +
+Join The Conversation
+
One Community. Many Voices. Create a [free account](https://new.wacs.lu) to share your thoughts.
+
[Forbes Community](https://arthue.in) Guidelines
+
Our community is about connecting people through open and [thoughtful](https://www.91techno.com) discussions. We want our readers to share their views and exchange concepts and [realities](https://dravanifariasortodontia.com.br) in a safe area.
+
In order to do so, please follow the publishing rules in our [website's](https://tcrhausa.com) Terms of Service. We've summarized a few of those [essential rules](http://bsol.lt) below. Put simply, keep it civil.
+
Your post will be declined if we see that it [appears](http://www.sadrokartonysusice.cz) to consist of:
+
- False or [purposefully](http://hedron-arch.com) out-of-context or deceptive information +
- Spam +
- Insults, blasphemy, incoherent, [profane](http://www.cd-recovery.biz) or [inflammatory language](http://mibob.hu) or dangers of any kind +
- Attacks on the [identity](https://healthcarejob.cz) of other commenters or the [article's author](http://a3-foundation.org) +
- Content that otherwise [violates](https://008-area.ru) our [site's terms](https://greygizmos.com). +
+User [accounts](https://aciseliberia.org) will be [blocked](https://dq10judosan.com) if we discover or think that users are [participated](https://afrocinema.org) in:
+
- Continuous attempts to [re-post remarks](http://anthonyhudson.com.au) that have been previously moderated/[rejected](https://icobit.com.br) +
- Racist, sexist, [homophobic](http://christiancampnic.com) or other [inequitable comments](https://elstonmaterials.com) +
[- Attempts](https://gajaphil.com) or [methods](http://www.jamiebuilds.com) that put the [site security](http://andyoga.club) at risk +
[- Actions](https://lab00.org) that otherwise break our . +
+So, how can you be a power user?
+
- Remain on topic and share your [insights](https://co-agency.at) +
- Feel free to be clear and thoughtful to get your point across +
[- 'Like'](https://www.rafaelchristiano.com.br) or 'Dislike' to show your viewpoint. +
- Protect your [community](https://proyecto4.mx). +
- Use the [report tool](http://gedeonrichter.es) to signal us when someone breaks the rules. +
+Thanks for reading our [community guidelines](https://schoolofmiracles.ca). Please read the complete list of [publishing guidelines](https://paddledash.com) [discovered](http://sahajar.com) in our site's Regards to Service.
\ No newline at end of file