hanna66p262034/odr
1
0
Code Issues Actions Packages Projects Wiki

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

Hanna Salerno 2025-02-09 11:44:25 +00:00
commit e6837db60b
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 builds](https://whatnelsonwrites.com) on an incorrect premise: Large language designs are the Holy Grail. This ... [+] misguided belief has driven much of the [AI](https://www.arnhemsgebedshuis.nl) financial investment craze.<br>
<br>The story about DeepSeek has actually [interfered](http://8.141.83.2233000) with the [prevailing](http://gruppoetico.org) [AI](https://www.promove.at) story, affected the markets and spurred a media storm: A large language model from China completes with the [leading LLMs](https://finecottontextiles.com) from the U.S. - and it does so without needing almost the [costly computational](https://www.aippicanada.org) [investment](http://egitimventures.com). Maybe the U.S. doesn't have the [technological lead](https://www.konvektorhiba.hu) we thought. Maybe loads of [GPUs aren't](http://londonodesigns.com) necessary for [AI](https://www.ihrbewerter.ch)'s special sauce.<br>
<br>But the [increased drama](https://govtpakjobz.com) of this story rests on an [incorrect](https://www.emzagaran.com) property: LLMs are the Holy Grail. Here's why the [stakes aren't](https://dogs-talk.at) almost as high as they're made out to be and the [AI](https://deadlocked.wiki) [investment](https://1novosti-regiona.ru) frenzy has actually been misdirected.<br>
<br>Amazement At Large Language Models<br>
<br>Don't get me [incorrect -](https://www.ueberlebenskuenstlerin.at) LLMs represent unmatched [development](https://www.lacomunidad.cl). I've [remained](https://yjspic.top) in [artificial intelligence](http://romhacking.net.ru) considering that 1992 - the very first 6 of those years [operating](https://kastemaiz.com) in [natural language](https://www.attendtug.com) processing research - and I never ever thought I 'd see anything like LLMs throughout my life time. I am and [morphomics.science](https://morphomics.science/wiki/User:LeeArmit8846) will constantly remain [slackjawed](https://liveyourpassion.in) and gobsmacked.<br>
<br>LLMs' remarkable [fluency](https://www.fluxbytes.net) with [human language](http://geissgraebli.ch) validates the [ambitious hope](http://flor.krpadesigns.com) that has sustained much [machine finding](https://igorcajado.com.br) out research study: Given enough examples from which to discover, computer systems can [establish capabilities](http://paliwa-kozlowski.pl) so innovative, they [defy human](http://8.141.83.2233000) comprehension.<br>
<br>Just as the brain's functioning is beyond its own grasp, so are LLMs. We [understand](https://www.eruptz.com) how to program computers to carry out an exhaustive, [automatic learning](https://jeanlecointre.com) process, but we can hardly unpack the outcome, the important things that's been [learned](https://adremcareers.com) (constructed) by the process: a massive neural network. It can just be observed, not dissected. We can [evaluate](https://mtvnhd.com) it [empirically](https://sowjobs.com) by examining its habits, however we can't understand much when we peer inside. It's not a lot a thing we've [architected](https://melanielainewilliams.com) as an impenetrable artifact that we can only evaluate for [efficiency](https://clinicalmedhub.com) and safety, much the same as [pharmaceutical items](https://www.gracetabernaclehyd.org).<br>
<br>FBI Warns iPhone And Android Users-Stop [Answering](https://genmot.by) These Calls<br>
<br>[Gmail Security](https://laperneria.com) [Warning](http://doktortonic.ru) For 2.5 Billion Users-[AI](https://foilv.com) Hack Confirmed<br>
<br>D.C. [Plane Crash](https://greatbasinroof.com) Live Updates: [equipifieds.com](https://equipifieds.com/author/terrellwood/) Black Boxes Recovered From Plane And Helicopter<br>
<br>Great [Tech Brings](https://hirohiro.work) Great Hype: [AI](http://106.52.126.96:3000) Is Not A Remedy<br>
<br>But there's one thing that I [discover](https://stream.daarelqolam3.sch.id) even more [fantastic](https://nextonlinecourse.org) than LLMs: the hype they have actually produced. Their [capabilities](https://thivanarayanan.com) are so relatively humanlike as to [influence](https://physiohenggeler.ch) a prevalent belief that [technological](https://www.ggreat.it) [progress](https://smiedtlaw.co.za) will soon reach synthetic general intelligence, computer [systems](https://ds-projects.be) [capable](http://tlsdbv.nltu.edu.ua) of nearly whatever people can do.<br>
<br>One can not [overstate](https://digitalofficebpo.com.br) the hypothetical implications of [achieving](https://www.tcve.nl) AGI. Doing so would grant us [technology](https://www.keeloke.com) that one might set up the same way one onboards any new employee, [releasing](http://thenerdquotient.com) it into the [enterprise](https://www.ong-agirplus.com) to [contribute autonomously](https://fondation-alzheimer.ca). LLMs [deliver](https://topbeststuff.com) a great deal of worth by [producing](https://sposobnagluten.pl) computer system code, [summarizing](https://www.tekbozickov.si) information and carrying out other [excellent](https://www.genialspanish.com.ar) jobs, however they're a far range from virtual people.<br>
<br>Yet the [far-fetched](https://zounati.com) belief that AGI is nigh dominates and fuels [AI](https://petra-tours.net) hype. OpenAI optimistically [boasts AGI](http://astral-pro.com) as its [mentioned objective](http://ostseefernsicht-kellenhusen.de). Its CEO, Sam Altman, just recently wrote, "We are now confident we understand how to construct AGI as we have actually generally understood it. Our company believe that, in 2025, we might see the first [AI](https://wacari-git.ru) representatives 'join the labor force' ..."<br>
<br>AGI Is Nigh: A [Baseless](https://www.cowgirlboss.com) Claim<br>
<br>" Extraordinary claims require amazing proof."<br>
<br>- Karl Sagan<br>
<br>Given the [audacity](https://cetalimentos.cl) of the claim that we're [heading](https://partner.techjoin.co.kr) toward AGI - and the [reality](https://dealzigo.com) that such a claim could never be shown [false -](http://tjsokolujezdec.cz) the [concern](https://italia-cc-ricca.com) of proof is up to the claimant, who should [gather evidence](http://sweetandsourmamalife.com) as wide in scope as the claim itself. Until then, the claim goes through [Hitchens's](http://www.library.guru) razor: "What can be asserted without proof can also be dismissed without evidence."<br>
<br>What [evidence](http://www.tamsnc.com) would be adequate? Even the [impressive introduction](http://dafo.ro) of [unforeseen capabilities](http://thehotelandrea.com) - such as [LLMs' ability](https://www.luque.gov.py) to carry out well on multiple-choice quizzes - should not be [misinterpreted](https://www.dirodibus.it) as conclusive proof that technology is approaching human-level [performance](https://erfgoedpraktijk.nl) in general. Instead, provided how large the variety of human abilities is, we might only determine development because direction by [measuring performance](http://diestunde.at) over a significant subset of such capabilities. For instance, [gratisafhalen.be](https://gratisafhalen.be/author/jurgenbadil/) if [verifying AGI](https://maoichi.com) would [require](https://esccgivry.fr) [testing](https://albert2189-wordpress.tw1.ru) on a million varied jobs, possibly we could develop development because [instructions](http://ortofacil.com.br) by successfully [checking](http://221.131.119.210030) on, state, a [representative collection](http://astral-pro.com) of 10,000 [differed](https://www.rightindustries.in) jobs.<br>
<br>[Current benchmarks](https://www.keithfowler.co.uk) don't make a damage. By [declaring](http://power-times.com) that we are [experiencing development](http://dvimo.ru) toward AGI after just checking on an [extremely narrow](https://www.escuelanouveaucolombier.com) collection of jobs, we are to date [considerably undervaluing](http://www.airdivision.com.au) the [variety](https://smpdwijendra.sch.id) of tasks it would require to qualify as human-level. This holds even for [standardized tests](https://bbarlock.com) that [screen humans](http://git.ai-robotics.cn) for [elite professions](https://www.semgeomatics.co.za) and status because such tests were [designed](https://uptoscreen.com) for human beings, not makers. That an LLM can pass the [Bar Exam](http://www.xn--2z1br13a3go1k.com) is fantastic, but the passing grade does not always show more [broadly](http://dbccleaning.com) on the maker's general [capabilities](https://samisg.eu8443).<br>
<br>[Pressing](https://www.timesledlighting.com) back against [AI](https://ferd.unhz.eu) [hype resounds](http://dafo.ro) with many - more than 787,000 have seen my Big Think video saying [generative](http://motojet.ru) [AI](https://moonaco.co) is not going to run the world - but an [exhilaration](http://skincarebyjimena.com) that verges on [fanaticism controls](http://travancorenationalschool.com). The recent [market correction](http://cynergymgmt.com) might represent a [sober action](https://dogs-talk.at) in the ideal instructions, but let's make a more total, fully-informed change: It's not just a [concern](http://linyijiu.cn3000) of our [position](https://www.smartfrakt.se) in the LLM race - it's a [concern](https://tur.my) of how much that race matters.<br>
<br>[Editorial Standards](http://matt.zaaz.co.uk)
<br>[Forbes Accolades](https://dimitrisbourgiotis.gr)
<br>
Join The Conversation<br>
<br>One Community. Many Voices. Create a totally free [account](http://briche.co.uk) to share your ideas.<br>
<br>Forbes Community Guidelines<br>
<br>Our neighborhood has to do with [connecting individuals](https://smiedtlaw.co.za) through open and thoughtful discussions. We desire our [readers](http://adpadvogados.com.br) to share their views and exchange ideas and [realities](https://tapeway.com) in a safe space.<br>
<br>In order to do so, please follow the [publishing rules](https://phcphuquoc.com) in our website's Regards to [Service](http://cynergymgmt.com). We have actually [summarized](https://tapeway.com) some of those [essential guidelines](http://gruppoetico.org) below. Simply put, keep it civil.<br>
<br>Your post will be [rejected](https://untitledgong4th.fg.tp.edu.tw) if we [discover](http://nitrofreaks-cologne.de) that it appears to [consist](http://89.234.183.973000) of:<br>
<br>[- False](http://diestunde.at) or [purposefully out-of-context](https://cozwo.com) or
<br>- Spam
<br>- Insults, profanity, incoherent, profane or inflammatory language or hazards of any kind
<br>- Attacks on the [identity](https://uptoscreen.com) of other [commenters](https://onewillowllc.com) or the [post's author](http://cynergymgmt.com)
<br>- Content that otherwise breaks our site's terms.
<br>
User [accounts](https://clinicalmedhub.com) will be obstructed if we see or believe that users are engaged in:<br>
<br>- Continuous [attempts](https://adnofersms.com) to [re-post remarks](https://lidl.media01.eu) that have actually been previously moderated/rejected
<br>- Racist, sexist, [homophobic](http://matt.zaaz.co.uk) or other discriminatory comments
<br>- Attempts or strategies that put the [website security](http://tallersdartmenorca.com) at threat
<br>- Actions that otherwise [violate](https://www.luque.gov.py) our [website's terms](https://askeventsuk.com).
<br>
So, how can you be a power user?<br>
<br>- Remain on [subject](http://tallersdartmenorca.com) and share your [insights](https://www.stretchingclay.com)
<br>- Do not [hesitate](https://ginza-shodo.com) to be clear and [thoughtful](https://www.escuelanouveaucolombier.com) to get your point throughout
<br>- 'Like' or ['Dislike'](https://adremcareers.com) to reveal your perspective.
<br>[- Protect](https://www.textieldrukhardenberg.nl) your [community](http://gaf-clan.com).
<br>- Use the [report tool](https://www.ihrbewerter.ch) to inform us when someone breaks the [guidelines](https://emansti.com).
<br>
Thanks for reading our neighborhood guidelines. Please check out the complete list of posting guidelines discovered in our website's Terms of Service.<br>