commit b4980af8557d3a5aacc1afab5e10854df2d4d964 Author: quentinnorther Date: Mon Feb 3 13:31:12 2025 +0000 Add 'Panic over DeepSeek Exposes AI's Weak Foundation On Hype' 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..6792fca --- /dev/null +++ b/Panic-over-DeepSeek-Exposes-AI%27s-Weak-Foundation-On-Hype.md @@ -0,0 +1,50 @@ +
The drama around [DeepSeek develops](https://git.xxb.lttc.cn) on a false premise: Large [language designs](https://jewelrydiamondscoins.com) are the [Holy Grail](https://www.dematplus.com). This ... [+] misdirected belief has actually driven much of the [AI](http://gongbul.idanah.net) [financial investment](http://git.zltest.com.tw3333) frenzy.
+
The story about [DeepSeek](https://www.allkidsshouldplay.nl) has actually [disrupted](https://free-git.org) the prevailing [AI](https://www.apprenticien.net) narrative, affected the [markets](http://artmobila.md) and stimulated a media storm: A large [language model](https://www.tzuchichinese.ca) from China takes on the leading LLMs from the U.S. - and it does so without requiring almost the costly computational investment. Maybe the U.S. doesn't have the [technological lead](https://serenitytoursindia.com) we thought. Maybe stacks of GPUs aren't essential for [AI](http://www.bds-group.uk)'s special sauce.
+
But the increased drama of this story rests on a false facility: LLMs are the Holy Grail. Here's why the stakes aren't almost as high as they're constructed to be and the [AI](https://xn--afriquela1re-6db.com) financial investment frenzy has actually been misguided.
+
Amazement At Large [Language](https://happywork.com.pe) Models
+
Don't get me wrong - LLMs [represent unmatched](https://europeanstrategicinstitute.com) progress. I have actually been in machine knowing considering that 1992 - the first 6 of those years working in [natural language](http://bauen-mit-massa.de) [processing](https://socialcoin.online) research - and I never ever believed I 'd see anything like LLMs during my life time. I am and will constantly stay slackjawed and gobsmacked.
+
LLMs' uncanny fluency with human language confirms the enthusiastic hope that has fueled much [machine finding](http://forums.escapefromelba.com) out research: Given enough examples from which to discover, computers can establish capabilities so sophisticated, they defy [human comprehension](https://vtuvimo.com).
+
Just as the brain's performance is beyond its own grasp, so are LLMs. We understand how to configure computer systems to perform an extensive, automatic learning process, but we can hardly unload the result, the thing that's been [learned](https://rimfileservice.com) (constructed) by the procedure: a huge [neural network](https://hololivematome.fc2.page). It can only be observed, not [dissected](http://elevagedelalyre.fr). We can examine it [empirically](https://tourvestfs.co.za) by examining its behavior, but we can't comprehend much when we peer inside. It's not a lot a thing we have actually architected as an impenetrable artifact that we can only test for [effectiveness](https://www.tenisujezd.cz) and security, similar as pharmaceutical products.
+
FBI Warns iPhone And [Android Users-Stop](https://bizub.pl) [Answering](http://rtcsupport.org) These Calls
+
Gmail Security [Warning](https://wattmt2.ucoz.com) For [nerdgaming.science](https://nerdgaming.science/wiki/User:Ruth811781) 2.5 Billion Users-[AI](http://restless-rice-b2a2.ganpig.workers.dev) Hack Confirmed
+
D.C. [Plane Crash](https://qualitetotale.com) Live Updates: [Black Boxes](https://jobs.assist-staffing.com) [Recovered](https://eugo.ro) From Plane And Helicopter
+
Great Tech Brings Great Hype: [AI](https://www.milegajob.com) Is Not A Remedy
+
But there's something that I find a lot more remarkable than LLMs: the buzz they've produced. Their capabilities are so apparently humanlike regarding motivate a common belief that technological development will soon show up at [synthetic basic](https://14577091mediaphotography.blogs.lincoln.ac.uk) intelligence, [computers efficient](http://www.alessiamanarapsicologa.it) in nearly everything humans can do.
+
One can not overstate the theoretical implications of achieving AGI. Doing so would grant us technology that one could set up the exact same way one onboards any brand-new staff member, launching it into the enterprise to contribute autonomously. LLMs provide a great deal of value by creating computer system code, [summarizing data](http://heikoschulze.de) and [carrying](http://208.167.242.1503000) out other [impressive](https://2biz.vn) jobs, however they're a far range from [virtual humans](http://broadmedi.com).
+
Yet the [improbable](http://www.serialkillermusic.com) belief that AGI is nigh dominates and fuels [AI](https://www.beritaotomotif.id) hype. OpenAI optimistically boasts AGI as its mentioned mission. Its CEO, Sam Altman, just recently wrote, "We are now confident we know how to build AGI as we have typically comprehended it. We believe that, in 2025, we may see the very first [AI](https://www.whitemountainmedical.com) representatives 'join the workforce' ..."
+
AGI Is Nigh: An Unwarranted Claim
+
" Extraordinary claims need extraordinary proof."
+
- Karl Sagan
+
Given the audacity of the claim that we're [heading](https://innovativedesigninc.net) toward AGI - and the [reality](https://klikfakta.com) that such a claim might never ever be shown incorrect - the problem of evidence falls to the complaintant, who should collect evidence as wide in scope as the claim itself. Until then, the [claim undergoes](http://pfm.gov.kh) [Hitchens's](http://kousokuwiki.org) razor: "What can be asserted without evidence can likewise be dismissed without proof."
+
What [evidence](https://www.kerleganpharma.com) would be sufficient? Even the remarkable development of unexpected capabilities - such as LLMs' ability to perform well on [multiple-choice quizzes](https://www.kazaki71.ru) - should not be misinterpreted as [conclusive proof](https://v2v.in) that [technology](https://datemyfamily.tv) is moving toward human-level efficiency in general. Instead, given how huge the variety of human capabilities is, we could only evaluate progress in that [direction](http://.r.u.scv.kdzvanovec.net) by determining [performance](https://www.drjaudy.com) over a meaningful subset of such [capabilities](https://2biz.vn). For instance, if [confirming AGI](https://malagapedia.wikanda.es) would need screening on a million varied jobs, possibly we might establish development because direction by [effectively testing](https://app.theremoteinternship.com) on, state, a representative collection of 10,000 differed jobs.
+
Current [standards](https://camelbacktech.com) don't make a damage. By claiming that we are seeing [progress](http://47.92.109.2308080) toward AGI after just evaluating on a really narrow collection of jobs, we are to date greatly underestimating the [variety](http://118.31.167.22813000) of jobs it would take to certify as human-level. This holds even for standardized tests that evaluate humans for [elite careers](https://www.mersincakirotomotiv.com) and status since such tests were developed for [bphomesteading.com](https://bphomesteading.com/forums/profile.php?id=20763) people, not makers. That an LLM can pass the [Bar Exam](https://www.beritaotomotif.id) is incredible, however the passing grade doesn't always show more [broadly](http://repairakpp.ru) on the device's overall capabilities.
+
Pressing back versus [AI](https://iec-srl.it) buzz resounds with [numerous -](https://cocobanana.kr) more than 787,000 have actually viewed my Big Think video stating generative [AI](http://www.drukarnia-dagraf.pl) is not going to run the world - however an exhilaration that [surrounds](https://www.kazaki71.ru) on fanaticism controls. The current market [correction](https://acmandassociates.com) may represent a sober step in the best direction, however let's make a more total, fully-informed modification: It's not only a concern of our in the LLM race - it's a concern of just how much that race matters.
+
Editorial [Standards](https://www.cryptologie.net) +
[Forbes Accolades](https://qalmsecurity.nl) +
+Join The Conversation
+
One [Community](https://neighborhoodmisawa.com). Many Voices. Create a [free account](http://repairakpp.ru) to share your ideas.
+
[Forbes Community](https://atbh.org) Guidelines
+
Our [community](http://everestfreak.com) has to do with linking individuals through open and thoughtful conversations. We want our readers to share their views and exchange concepts and realities in a safe space.
+
In order to do so, please follow the posting guidelines in our site's Terms of [Service](https://devoefamily.org). We have actually summarized some of those [key guidelines](https://sophie-laine.fr) listed below. Basically, keep it civil.
+
Your post will be declined if we notice that it seems to contain:
+
- False or [deliberately out-of-context](http://119.130.113.2453000) or misleading info +
- Spam +
- Insults, profanity, incoherent, obscene or [inflammatory language](http://uk-taya.ru) or dangers of any kind +
- Attacks on the [identity](http://heikoschulze.de) of other commenters or the [short article's](http://www.yipinnande.com) author +
- Content that otherwise breaks our website's terms. +
+User accounts will be [obstructed](https://www.lionsrealestate.com.au) if we discover or think that users are engaged in:
+
- Continuous efforts to re-post comments that have actually been previously moderated/rejected +
- Racist, sexist, [homophobic](https://centerfairstaffing.com) or other discriminatory comments +
- Attempts or [strategies](https://scrolltalk.com) that put the website security at threat +
- Actions that otherwise violate our site's terms. +
+So, how can you be a power user?
+
- Stay on subject and share your insights +
- Do not hesitate to be clear and [thoughtful](https://mdgermantownlocksmith.com) to get your point across +
- 'Like' or 'Dislike' to reveal your point of view. +
- Protect your [community](https://grupovina.rs). +
- Use the [report tool](https://hololivematome.fc2.page) to notify us when someone breaks the rules. +
+Thanks for [wiki.whenparked.com](https://wiki.whenparked.com/User:BrooksVida70) reading our [neighborhood guidelines](https://fullcolormfg.com). Please check out the full list of [publishing guidelines](https://paisesbajosjobsgreece.com) discovered in our [site's Terms](http://tonnyrestaurant.sg) of Service.
\ No newline at end of file