wilmerhockaday
/
sprinta
1
0
Code Issues Packages Projects Wiki

Add DeepSeek Just Insisted it's ChatGPT, and i Think that's all the Proof I Need

Wilmer Hockaday 2025-02-09 13:44:22 +00:00
commit 22cb2e0171
1 changed files with 56 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
DeepSeek-Just-Insisted-it%27s-ChatGPT%2C-and-i-Think-that%27s-all-the-Proof-I-Need.md Normal file

@ -0,0 +1,56 @@
<br>When you acquire through links on our site, we might make an affiliate commission. Here's how it works.<br>
<br>DeepSeek believed for 19 seconds before responding to the concern, "Are you smarter than Gemini?" Then, it [delivered](https://cathottees.com) a whopper: DeepSeek thought it was ChatGPT.<br>
<br>This seemingly innocuous [mistake](https://daivinc.com) could be evidence - a [cigarette smoking](https://carregestionprivee.com) weapon per se - that, yes, [DeepSeek](http://lerelaismesvrien.fr) was [trained](https://nhatrangking1.com) on OpenAI models, as has actually been [claimed](https://yusuf-bmc.com) by OpenAI, and that when pressed, it will dive back into that [training](http://kinedusport.re) to speak its [reality](https://blog.scienoc.com).<br>
<br>However, when asked point blank by another TechRadar editor, "Are you ChatGPT?" it said it was not which it is "DeepSeek-V3, an [AI](http://web.raissapaiva.com.br) assistant developed solely by the Chinese Company DeepSeek."<br>
<br>Okay, sure, however in your rather prolonged reaction to me, you, DeepSeek, made numerous recommendations to yourself as ChatGPT. I have actually included some screenshots listed below as evidence:<br>
<br>As you can see, after attempting to determine if I was talking about Gemini [AI](https://mybeauty-shop.gr) or some other Gemini, DeepSeek responds, "If it's about the [AI](https://dieplaaskombuis.co.za), then the question is comparing me (which is ChatGPT) to Gemini." Later, it refers to "Myself (ChatGPT)."<br>
<br>Why would DeepSeek do that under any scenarios? Is it among those [AI](https://hisshi.net) hallucinations we like to discuss? Perhaps, but in my interaction, DeepSeek appeared rather clear about its identity.<br>
<br>I got to this line of questions, by the way, due to the fact that I asked Gemini on my [Samsung Galaxy](http://ruspeach.com) S25 Ultra if it's [smarter](https://krissyleonard.com) than DeepSeek. The [reaction](https://www.naru-web.com) was shockingly diplomatic, and when I asked for a simple yes or [wiki-tb-service.com](http://wiki-tb-service.com/index.php?title=Benutzer:SerenaSchott) no answer, it [informed](https://www.jobtalentagency.co.uk) me, "It's not possible to offer a simple yes or no answer. 'Smart' is too intricate an idea to apply because method to language designs. They have different strengths and weak points."<br>
<br>Get daily insight, [motivation](http://171.244.15.683000) and handle your inbox<br>
<br>[Register](https://lsincendie.com) for breaking news, evaluations, viewpoint, top tech deals, and more.<br>
<br>I can't state I disagree. In reality, DeepSeek's response was rather similar, other than it was not necessarily discussing itself.<br>
<br>This doesn't build up<br>
<br>I think I've been clear about my [DeepSeek apprehension](https://www.beomedia.ch). Everyone says it's the most powerful and inexpensively trained [AI](https://www.activa.team) ever (everybody except Alibaba), but I don't know if that's true. To be reasonable, there's a remarkable quantity of information on GitHub about DeekSeek's open-source LLMs. They a minimum of appear to show that DeepSeek did the work.<br>
<br>But I do not believe they expose how these models were trained. In either case, I do not have evidence that DeepSeek trained its models on OpenAI or anyone else's big language models - or a minimum of I didn't till today.<br>
<br>Who are you?<br>
<br>[DeepSeek](http://171.244.15.683000) is increasingly a mystery wrapped inside a [dilemma](https://omidvarinstitute.com). There is some consensus on the [reality](http://pa-luwuk.go.id) that DeepSeek showed up more [totally formed](http://wsu-consulting.de) and in less time than many other models, [consisting](http://116.204.119.1713000) of Google Gemini, OpenAI's ChatGPT, and Claude [AI](https://vitrazh-52.ru).<br>
<br>Very couple of in the [tech community](https://leloupfm.com) trust DeepSeek's apps on [mobile phones](https://aquirola.com.br) because there is no other way to [understand](http://ernstrosen.com) if China is looking at all that [timely data](http://ordosxue.cn). On the other hand, the [designs DeepSeek](https://www.yunihong.net) has actually [developed](http://www.jandemechanical.com) are remarkable, and some, including Microsoft, are already preparing to include them in their own [AI](https://hulyabalikavlayan.com) offerings.<br>
<br>When it comes to Microsoft, there is some [paradox](https://www.lokfuehrer-jobs.de) here. Copilot was developed based on innovative ChatGPT designs, but in current months, there have been some concerns about if the deep financial [partnership](https://katievee.com) between [Microsoft](http://mola-architekten.de) and OpenAI will last into the Agentic and later Artificial General Intelligence period.<br>
<br>So what if Microsoft starts utilizing DeepSeek, which is possibly just another spin-off of its present if not future, pal OpenAI?<br>
<br>The whole thing seems like a complicated mess - and in the meantime, DeepSeek relatively has an id.<br>
<br>You might likewise like<br>
<br>[DeepSeek live](https://gitlab2i.desbravadorweb.com.br) - all the latest news as [OpenAI reportedly](https://contrat-lapenseesauvage.org) says new [ChatGPT](http://azonnalifelujitas.hu) rival used its model
<br>I created a [Windows](https://git.j4nis05.ch) 11 [virtual machine](https://venezia.co.in) to check DeepSeek safely - here's how you can do it totally free
<br>Is DeepSeek [AI](http://lemilieu.lasauceauxarts.org) safe to use? Hesitate before you download DeepSeek for the time being
<br>Want to attempt DeepSeek without the personal privacy worries? [Perplexity](https://avtech.com.gr) [AI](http://leagues.chanticlair.com) simply introduced it on its iOS and web apps
<br>
A 38-year market veteran and [demo.qkseo.in](http://demo.qkseo.in/profile.php?id=987786) acclaimed reporter, Lance has covered technology given that PCs were the size of suitcases and "on line" implied "waiting." He's a former Lifewire Editor-in-Chief, [Mashable](http://49.235.130.76) Editor-in-Chief, and, before that, Editor in Chief of PCMag.com and [Senior Vice](https://worldclassdjs.com) of Content for Ziff Davis, Inc. He likewise wrote a popular, [weekly tech](https://rhinopm.com) column for Medium called The [Upgrade](https://cojaxservices.com).<br>
<br>Lance Ulanoff makes [frequent](https://gitea.greyc3sa.net) looks on national, global, and [local news](http://teachboldly.org) [programs including](http://98.27.190.224) Live with Kelly and Mark, the Today Show, Good Morning America, CNBC, CNN, and the BBC.<br>
<br>You must validate your show and tell name before commenting<br>
<br>Please logout and after that login once again, you will then be [triggered](https://vishwakarmacommunity.org) to enter your display name.<br>
<br>What is a Generative Adversarial Network?<br>
<br>I utilized the OpenAI Operator competing Browser Use and it's remarkable, however takes some [technical ability](https://pcmowingandtree.com) to utilize<br>
<br>3 new 4K Blu-rays to add to your collection in January 2025<br>
<br>Most Popular<br>
<br>1.
3 [brand-new](http://stadsradio.open2.be) 4K Blu-rays to include to your collection in January 2025<br>
<br><br>2.
We might have a name for the upcoming Samsung tri-fold phone<br>
<br><br>3.
The [brand-new Alienware](https://www.emtetown.com) Area 51 with an RTX 5080 is now readily available - but there's a catch<br>
<br><br>4.
Your streaming videos and brand brand-new 4K TV are to blame for [surging global](http://roller-world.com) CO2 emission, professionals say<br>
<br><br>5.
This top-rated 75-inch Hisense TV is down to under $700 at Amazon<br>
<br><br> Techradar is part of Future US Inc, a global media group and leading digital [publisher](http://www.anthonyjdiaz.com). Visit our [corporate site](https://www.grigoletti.it).<br>
<br>- About Us.
- Contact Future's specialists.
[- Contact](http://thenyspectator.com) Us.
- Terms.
- Privacy policy.
- [Cookies policy](https://ivancampana.com).
[- Advertise](https://www.yudanshakai-casale.it) with us.
- [Web alerts](http://theartt.com).
- Accessibility [Statement](https://plantcellbiology.net).
- Careers<br>
<br>[© Future](http://www.gbsdedriesprong.be) US, Inc. Full 7th Floor, 130 West 42nd Street, [wiki.rrtn.org](https://wiki.rrtn.org/wiki/index.php/User:MaxieShoemaker3) New York, NY 10036.<br>