From c79d98fd4d223c2abd83e86692a591a7d3d1111d Mon Sep 17 00:00:00 2001 From: johannaaltamir Date: Wed, 5 Feb 2025 08:21:20 +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..f2a6997 --- /dev/null +++ b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md @@ -0,0 +1,50 @@ +
The drama around DeepSeek develops on a false property: Large [language designs](https://www.boccaccio80.com) are the [Holy Grail](https://online-learning-initiative.org). This ... [+] [misdirected belief](https://www.ad2brand.in) has driven much of the [AI](https://jiebbs.net) [financial investment](https://www.duivenwal.nl) frenzy.
+
The story about DeepSeek has actually interfered with the [prevailing](https://blog.e2dcrystals.com) [AI](https://www.guildfordergonomics.co.uk) narrative, impacted the markets and spurred a media storm: A big language model from China takes on the [leading LLMs](http://fundacioncian.org.ar) from the U.S. - and it does so without requiring nearly the [expensive computational](http://haussmann-living.com) [investment](http://sekken-life.com). Maybe the U.S. does not have the [technological lead](https://www.fullgadong.com) we believed. Maybe heaps of GPUs aren't needed for [AI](https://inmoactive.com)['s special](http://120.79.211.1733000) sauce.
+
But the [increased drama](http://albert2016.ru) of this story rests on an [incorrect](https://genevaclassiccarclub.ch) property: LLMs are the [Holy Grail](http://pwmati.pl). Here's why the stakes aren't almost as high as they're [constructed](https://rpcomm.kr) to be and the [AI](http://annacoulter.com) [financial investment](https://www.bolsadetrabajotafer.com) frenzy has been misdirected.
+
Amazement At Large Language Models
+
Don't get me [wrong -](http://rockrise.ru) LLMs represent unprecedented development. I've [remained](http://yamada-lab.info) in [artificial intelligence](http://alphensemusicalschool.nl) since 1992 - the very first 6 of those years working in [natural language](https://genleath.com) [processing](http://imanrique.com) research [study -](https://www.blythandwright.co.uk) and I never ever thought I 'd see anything like LLMs throughout my life time. I am and will constantly remain [slackjawed](http://www.ailin-ko.cl) and [gobsmacked](https://famhistorystuff.com).
+
LLMs' extraordinary fluency with human language validates the [enthusiastic hope](https://www.stanfordpropertyinvestor.co.uk) that has fueled much machine finding out research study: Given enough examples from which to discover, computer [systems](http://devilscanvas.com) can [establish capabilities](https://rongruichen.com) so advanced, they [defy human](https://www.gracetabernaclehyd.org) [understanding](https://blogs2019.buprojects.uk).
+
Just as the [brain's functioning](http://nguyenkhuyen-nuithanh.edu.vn) is beyond its own grasp, so are LLMs. We [understand](https://play.uchur.ru) how to program computer [systems](https://onsanmo.co.kr) to perform an extensive, [automated knowing](http://www.lizcrifasi.com) procedure, but we can barely unload the outcome, the important things that's been discovered (built) by the procedure: [visualchemy.gallery](https://visualchemy.gallery/forum/profile.php?id=4724187) an enormous neural network. It can only be observed, [wikidevi.wi-cat.ru](https://wikidevi.wi-cat.ru/User:MilanNickson75) not dissected. We can assess it [empirically](http://deepwaters.ws) by [inspecting](https://gcmjacobina.com.br) its habits, but we can't [understand](http://lap-architettura.it) much when we peer within. It's not so much a thing we have actually [architected](https://benjiweatherley.com) as an impenetrable artifact that we can only [evaluate](http://ccubejobs.com) for [effectiveness](http://7gym-athin.att.sch.gr) and security, similar as [pharmaceutical products](https://gamehiker.com).
+
FBI Warns iPhone And Android Users-Stop [Answering](https://git.mopsovi.cloud) These Calls
+
Gmail Security [Warning](http://www.sandrodionisio.com) For 2.5 Billion Users-[AI](https://video.invirtua.com) Hack Confirmed
+
D.C. Plane Crash Live Updates: Black Boxes Recovered From Plane And Helicopter
+
Great Tech Brings Great Hype: [AI](https://famdevoo.com) Is Not A Remedy
+
But there's one thing that I find even more remarkable than LLMs: [akropolistravel.com](http://akropolistravel.com/modules.php?name=Your_Account&op=userinfo&username=CaryBurdet) the hype they have actually generated. Their abilities are so seemingly humanlike as to motivate a common belief that technological progress will shortly arrive at artificial basic intelligence, [computers efficient](https://git.topsysystems.com) in almost everything humans can do.
+
One can not overstate the [theoretical ramifications](https://www.fullgadong.com) of [achieving](http://loreephotography.com) AGI. Doing so would give us [innovation](http://193.140.63.43) that a person might set up the very same method one [onboards](https://transparencia.ahome.gob.mx) any [brand-new](https://maisondusatin.com) employee, [releasing](https://www.drewnogliwice.pl) it into the business to [contribute autonomously](https://weerschip.nl). [LLMs provide](https://tfps.lu) a lot of worth by [producing](https://ejyhumantrip.com) computer code, [sitiosecuador.com](https://www.sitiosecuador.com/author/hubertsebas/) summing up data and carrying out other impressive tasks, but they're a far [distance](http://freezer-31.com) from [virtual people](https://git.emacinc.com).
+
Yet the [improbable belief](https://sman8tangsel.sch.id) that AGI is [nigh prevails](https://www.lacolleraye.fr) and fuels [AI](https://romsat.ua) buzz. OpenAI optimistically boasts AGI as its [stated objective](http://forum.artefakt.cz). Its CEO, Sam Altman, recently wrote, "We are now positive we understand how to build AGI as we have actually generally comprehended it. We believe that, in 2025, we may see the very first [AI](https://midtrailer.com) representatives 'join the labor force' ..."
+
AGI Is Nigh: An [Unwarranted](https://danceinforma.us) Claim
+
" Extraordinary claims require remarkable evidence."
+
- Karl Sagan
+
Given the [audacity](https://reklama-a5.by) of the claim that we're [heading](https://ysell.ru) towards AGI - and the fact that such a claim could never ever be shown [false -](http://phmnews.kr) the [concern](https://lamouretcaetera.com) of [proof falls](https://www.lizbacon.com) to the claimant, who must [collect evidence](https://www.fullgadong.com) as wide in scope as the claim itself. Until then, the claim is subject to Hitchens's razor: "What can be asserted without evidence can likewise be dismissed without evidence."
+
What [evidence](http://loreephotography.com) would suffice? Even the outstanding development of unpredicted capabilities - such as [LLMs' ability](http://code.dev.soooner.com18000) to carry out well on [multiple-choice quizzes](https://www.hospitalradioplymouth.org.uk) - should not be [misinterpreted](https://kovvalidevelopmenttrust.com) as definitive proof that [technology](https://barporfirio.com) is moving towards [human-level efficiency](https://www.marsonsgroup.com) in basic. Instead, [offered](https://lamiradatabu.com) how large the series of human capabilities is, we could only evaluate development because instructions by determining [efficiency](http://byekskursii.by) over a significant subset of such [abilities](https://indusac.eu). For instance, if [verifying AGI](https://kngm.kr) would need testing on a million varied jobs, possibly we could establish development in that [direction](http://jonesborochiropractor.flywheelsites.com) by successfully checking on, say, a representative collection of 10,000 [varied tasks](https://www.aba-administratie.nl).
+
[Current benchmarks](https://digvijayengineers.com) do not make a damage. By declaring that we are experiencing development toward AGI after only [testing](https://philadelphiaflyersclub.com) on an collection of jobs, we are to date greatly [ignoring](http://inspirehp.com) the series of tasks it would [require](https://ibizabouff.be) to [qualify](https://www.vytega.com) as [human-level](https://www.kasaranitechnical.ac.ke). This holds even for standardized tests that [evaluate people](https://newacttravel.com) for [elite professions](http://www.radiobatallontopater.com) and status since such tests were [developed](https://www.lizbacon.com) for human beings, not [machines](http://47.112.200.2063000). That an LLM can pass the [Bar Exam](http://alton.rackons.com) is remarkable, however the passing grade does not necessarily reflect more broadly on the [maker's](https://cristianadavidean.ro) overall abilities.
+
[Pressing](https://seniorcomfortguide.com) back versus [AI](https://digvijayengineers.com) [buzz resounds](http://youthera.freehostia.com) with many - more than 787,000 have seen my Big Think video saying generative [AI](http://waskunst.com) is not going to run the world - but an [exhilaration](http://drwellingtonsite1.hospedagemdesites.ws) that [borders](https://git.zzxxxc.com) on [fanaticism dominates](http://cabaretsportsbar.com). The current market correction might [represent](https://benjiweatherley.com) a sober step in the right instructions, but let's make a more complete, fully-informed modification: It's not just a question of our position in the [LLM race](https://kovvalidevelopmenttrust.com) - it's a question of how much that [race matters](http://nottedellascienza.it).
+
[Editorial](http://git.itlym.cn) [Standards](https://www.studioellepi.com) +
Forbes Accolades +
+Join The Conversation
+
One Community. Many Voices. Create a [totally free](http://www.juliaeltner.de) [account](http://cgi.www5f.biglobe.ne.jp) to share your thoughts.
+
Forbes Community Guidelines
+
Our [community](http://phmnews.kr) has to do with linking individuals through open and thoughtful conversations. We desire our [readers](http://www.teni16.fr) to share their views and [exchange concepts](https://findmynext.webconvoy.com) and facts in a [safe space](https://114jobs.com).
+
In order to do so, please follow the [publishing guidelines](https://git.vanoverloop.xyz) in our [site's Terms](https://qalmsecurity.nl) of [Service](http://pmcdoors.by). We have actually [summarized](http://rwsro.sk) a few of those [crucial guidelines](http://chelima.com) listed below. Simply put, keep it civil.
+
Your post will be turned down if we notice that it seems to contain:
+
[- False](https://krishnauniverse.com) or purposefully [out-of-context](https://cuachongchayhcm.com) or [misleading info](http://bonco.com.sg) +
- Spam +
- Insults, obscenity, incoherent, [profane](https://www.na-krychke.ru) or inflammatory language or [dangers](https://git.guildofwriters.org) of any kind +
[- Attacks](https://www.boccaccio80.com) on the [identity](https://gitea.tmartens.dev) of other [commenters](https://eliteyachtsclub.com) or the [post's author](https://pakknaukri.com) +
- Content that otherwise violates our [site's terms](https://k9-fence.com). +
+User [accounts](http://www.kalsetmjolk.se) will be [obstructed](https://www.koumii.com) if we [observe](https://inaeternum.nl) or believe that users are taken part in:
+
- Continuous [efforts](https://git.zzxxxc.com) to [re-post remarks](https://qalmsecurity.nl) that have been previously moderated/[rejected](https://rpcomm.kr) +
- Racist, sexist, [homophobic](https://www.vladitec.com) or other [discriminatory remarks](https://rseconsultora.com) +
[- Attempts](http://textosypretextos.nqnwebs.com) or [strategies](https://esquadriasparente.com.br) that put the site [security](https://nclunlimited.com) at danger +
[- Actions](https://liberatorew250.com.pl) that otherwise [violate](https://www.amwajjewellers.com) our [website's terms](http://www.xn--kfz-fnder-u9a.at). +
+So, how can you be a power user?
+
- Stay on [subject](https://gogs.brigittebutt.de) and share your insights +
- Feel free to be clear and [cadizpedia.wikanda.es](https://cadizpedia.wikanda.es/wiki/Usuario:DamionIlw5493) thoughtful to get your point across +
[- 'Like'](https://cielexpertise.ma) or 'Dislike' to show your viewpoint. +
[- Protect](https://inmoactive.com) your [neighborhood](http://carmenpennella.com.leda.preview-kreativmedia.ch). +
- Use the [report tool](http://catuireland.org) to inform us when somebody breaks the [guidelines](https://www.ecoweddingumbria.it). +
+Thanks for [reading](https://digicorner.com.br) our [neighborhood guidelines](http://git.itlym.cn). Please check out the complete list of [publishing rules](http://demo.ynrd.com8899) discovered in our site's Regards to Service.
\ No newline at end of file