Wednesday, February 12, 2025

Hırsız hukukçu sistemi

Hukukçular bok değil. Liseden sonra ÖSYM sınavda bok olamayıp ancak hukuk kazanmışlar. Maaş alma entrikaları hiç bitmez. Hırsız vesayet davası hakimleri. Kıskandıkları insanlara kendileri gibi beyinsiz devlet psikologlarla bir olup iftira ile vasi atayıp miras çalan, devletten maaş ayarlayan hırsız hakimler. Bin sülaleniz, tırun çovuğunuz hırsız damgalısınız. Hırsız vesayet davası hukukçuları. 

Thursday, February 6, 2025

Erenköy Hastane deki beyinsiz psikolog

2023 Ocak ayında, deri pantolonlu psikolog, senin kafana sıçayım. Senin beyinsiz kafana mı kalmış yüzde hesap, en önemli günlük matematik hesap yapmak. O beyinsiz kafanla birde rakam matematik ağzına alıp o hastanede bok biliyorsun elalemi mi kandırıyorsun. O Erenköy hastanenin bile seviyesini düşürdün. Geri zekalı beyinsiz, sen Einstein gelse o onun formülünü ona bile öğretmeye kalkacak beyinsiz manyaksın. O hastanenin değerine sıçma. Türkiye senin gibi kendini bilmezler yüzünden böyle geri. Sana mı kalmış uzun dönem plan program proje yapmak, doğuştan geri zekalı seni, haddini bilmez. Kıskanç. Liseden sonra ÖSYM sınavı nda altına sıçmışsın, Matematik Bilim gerçek sınav düşmanı seni, tuvalet kapısında sınav yaratıp gerçek sınavları aşağılamak amacın. Boksan 300 bin kişilik sınavda göster marifetlerini matematik bilimini. Beyinsiz Erenköy Hastane psikoloji bölüm çalışanı. Sen kenefte girdiğin sınavı değerli 300 bin kişilik yarışma sınavına eş tutmak istiyorsun. Ayrıca Türkiye de başkaları o senin yaptığın testi kimseye yapmadıklarını hatta birçok psikolog o testi kendileri başaramadıklarını, hastanede o test yapılan test değil dediler. Kıskanç kenef sınav meraklısı seni.
Geri zekalı seni. Karlıma şnsan diye alınacak beyin mi var sende. Geri zekalı, senin yaptığın o testi, yapamadığın testi, Lütfi Kırdar Şehir Hastanesi Psikoloji bölümü bilmedikleri yapamadıklarını en azından ikrar ettiler, geri zekalı. Kaç kişi ye yaptın, kaç sonuç, o dokumanların nerede. Dolandırıcı Erenköy Hastane çalışanı seni.
Beyinsiz hayvanlar bile hayvan olduğunu biliyor. Sen insan kılıklı beyinsiz dolandırıcı ve plagiarist sin. Erenköy Hastane nin yüz karasısın. O hastanenin aptal dolu çalışanla olduğu ispatısın. Yaptığın yanında kalmayacak, tazminat alınmalı senden.
Neyi ne amaçla yapıyorsun, hangi standarda dayanak alıyorsun? ÖSYM binasını da yıksan bari, kıskanç seni. Liseden sonra ÖSYM başarısızlığını ya o bşnayı yakmak, yada başarılı olanlara sakdırarak yapma planın suç. ÖSYM sınav sistemlerini yıkmak için başka ne bok yapıyorsun? Kıskanç Erenköy Hastane çalışanı seni.
Ayrıca hastane organize suç örgütlenme yapıp, kendi başvuru randevuya gitmiş, danışan birine hakaret etmiştir. Manyak Erenköy çalışanı, o gün orada bulunma nedeni sormadan, kişiye hakaret etmiştir.
Kişi sen matematik biliyorsan sınavın nerde demiş, Erenköy Hastane çalışanı pişkin gülmüş yetkinliğini ıspat eden belge ve fiil sunmamıştır. Hastane kapıyı kilitlemiş, kişiye işkence özgürlük kısıtlama yapmıştır. 
Erenköy Hastane çalışanı, seni kar kışda soğuk tuvalete tıkayıp, kapıdan sana adını orada neden olduğunu soracak bir senaryo da cevabını alacak birini arıyorum. Tuvalete tıkılmış olduğunu söylemeden cevabını bu ilan edeceğim. Tuvalette osuruk attığını kaydedip, şu soruya cevap osurdu, manyaktır diyeceğim. Gör iftira nedir. 
Senin örgütlenmiş suçun büyük ey Erenköy Hastane çalışanı. Gör bakalım Erenköy Hastane çalışanı işkence yapması ne.
O soğukta tuvalette olduğunu değil. Osurdu, sıçtı, şu soruyu sormuştum, manyaktır diyeceğim. İşkence altındaydı demeyeceğim tıpkı senin gibi. Bakalım adının osuruk, çiş olduğunu dediğimde ne diyeceksin. 
Ayrıca liseden sonra ÖSYM sınavı al dediğimi, o Erenköy Hastane çalışanın ise sadece gidip sıçtığını ve 300 bin kişilik sınavds bok olamadığını diyeceğim.
Sen boksan, sınavını o alanda başarını ispat et. Başka organizasyon ve kişileri umpasa düşürüp saldırarak iftira ile beyinsiz olduğunu saklamaya çalışma. Senin ne belgen var. İspat et. Sana orada hakaret ettiğimi bile dememişsin. Sana sen haddını bil bilmediğini bil, sınavın yok dediğimi dememişsin. Bok olsan ha evet onu ösym sınavda sordular, tanıyorum diyemedin bile. Pişkin pişkin beyinsizliğinden utanmadın bile. Resmi geçmişin geri zekalı çıktın dedim sana, pişkin sırıttın. Kedim bile daha anlardı ne hakaret ettim. Dolandırıcı sahteci seni. 
Denilenleri yalan eksik anlatmışsın.
Liseden sonra ÖSYM sınavı seçim sonucuna bok atmışsın yalancı Erenköy Hastane çalışanı. Hergün yapılsa seçim ülke korunamaz. Manyak Erenköy Hastane çalışanı.
Bu arada mesela bazı hastalar, mesela kiremitleri sayar, ezberler, ama on sene oku okur, gerçek matematik olan bir denklem ilgili harfli formülü ezberleyemez, öyle bir denklem ile matematik işlem yapamaz, yani sen Erenköy Hastane çalışanı gibi manyak beyni eksik olur.
Sana sen bunu bilmeyen beyinsizin tekisin dedim, bunu kimseye dememişsin. Yada o dediklerin zaten senden beter beyinsiz. 

Wednesday, February 5, 2025

AI in text citing errors

AI and Referenced Review Writing

ChatGPT and similar AI models are increasingly being used by students to complete assignments. Just a year ago, AI struggled with in-text citations. However, now, around twenty percent of in-text citation errors can reveal that an assignment was AI-generated, as many instructors recognize these patterns. 

Once this issue is fully resolved by better technology, it will become even harder to distinguish between AI-generated and student-written work by citing error pattern. 

And also proves student's inability, failure to audit an ai. I believe if a student fails to audit ai created citing, their grade need to be decreased. It is abuse of technology to deceit others that they know or abke to insert, match correct sources to correct claims. Newerdeless i catched that many students are dumb to correct citing if ai fails. I want students are at least able to audit ai citing errors. If those students gets full grade this would be injustice and abuse of valuable students who can even audit better than ai. Either do correct citing or do not. I also laughed on their stupidity that they all together did not understand others ai created erroric citing. It comes to a point that some did correct, they even failed to select better ai. 

I use in-text referencing error as marker of failure to understand a referenced paper, or even never read the paper but deceit instructor and other students as if they understood or audit their paper. This is an example thst the instructor ignores clear revealed plagiarism. Sure they are only student, nut at least warn that this is utmost important. The respect sure goes away for such audit deficit organization. 

I mean at least warning needed if they are only student. 

I decrease grade if no reference but find it plagiarism if uses erroric citing. I would even decrease the reputation of the organization for encouraging false behavior. 

Also such flaw even decrease the grade by points 3 or more over 10.That instructors overlook this flaw harms real reader knowledged audit students. 

If an instructor gives full grade to erroric citing full grade it even encourages plagiarism in students. Not correcting or at level informing citing error is sharing plagiarism if grade not dropped. 

Meanwhile i am not rven against if students uses pdf question tool to audit assignment, they even did not give 15 minutes more for ai itself checks citing. It improves better citing. On the other hand some students reads hours ıf they did correct citing. Damn such approaches. 

In turkish we say "bunlar namusunu kurtardı hocanın" what if some students were not excellent in-text citing audits. 

It is at least five hours to audit in-text referencing, and many students even fail after days of reading since not levelled to grasp ai cited reference. So some instructors unfairly give high grades to those who fail to audit ai citing error. Some students get low grade since absent listed sources, but if absence punished, then error need be punished more. Because it is ethically responsibility. Instructor duty is too heavy to catch those failed audit citing errors. 

Education grading has such infairness and distrupts trust in students grades. 

From my experience, I can identify approximately thirty percent of students using AI based on reference mistakes. When a student does not use references, I find it difficult to trust their work. On the other hand, I have observed AI generating error free in-text citations in many cases. Naturally, this capability will continue to improve over time.

Maybe in future this will be a duty of ai not human. 

As AI progresses, academic integrity concerns will become more significant. Institutions may need to develop more advanced detection methods or place greater emphasis on critical thinking and originality in assignments. AI is still worse than human in-text citing of a source. 

_____

If we want one day ai perfects in citing, we still need those auidit doing students to improve view of ai. I am fond of best ai, but ignored student in-text citing even do not improve ai development. We feed ai by asking best. 

Saturday, February 1, 2025

Siteler telefon onay kodu ister

. Siteler üyelik için ya onay kodu, ya telif başvuru formu ister. Türkiye de bir sitenin hukuki teknik özelliklerini hukuk açıdan bilen bilgisayar mühendisi bile yok. Dağbaşı ülkesi. Birisi telefonunu veremez. Çünki onay kodu verilmezse telefon kaydetmez İnternet site. 
Telefona gelen kodu ister. 
Bunu bilmeyen adam hukukçu olmuş. Türkiye hukukçular geri zekalı. 2017 de soru çalıntılı sınavla hukukçular sahtecilik yaptı.
Yani o site önce o kişinin elindeki telefona kod sms yollar sonra telefonu kayıt eder. Telefonu kod sormadan ilan etmişse, basittir, site kapatılır. Bu kadar hukuku, geri zekalı hakimler bilmediğini bilmiyor. Bilgisayar teknoloji bok biliyor hukukçular.

Yani, sahte çek alıp bankaya dalmak nasıl yanlışsa, öyle kanunsuz, bilgisayar kanuna uymayan siteye uymak suçtur. Site cybersecurity güvensiz ise, orada suç olacak fiilleri yapan tek celsede suçludur. Hukuk cahili hakimler eğitimli bilgisayar bilgili insanlara iftira vesayet davası açarsa, ilte Türkiye de bilgisayar anlayan kalmaz. 
Tabii suçludur, telefon kodu sormaması suç. 
Yani bu kadar basittir. Site bilgisayar kanunu beceremedi ise, o kişi nereden kanunsuz fiilinde bahane yaratıyor. Tek celsede suçtur o kişinin telefon kodu onayı olmayan siteyi bahane ile suç işlemesi. 
Hakimler 2017 den beri sıru çalıyor. TUS ve 2017 hukuk sınavları soru çalıntılı dır. Bok bilmeden soru çalıp o sınavı atlıyorlar. Bakıyorsun liseden sonra gerçek sınav da başarısızlar. Kanun bilen hakim yok. Bilgisayar teknoloji aptalı hukukçular akılılara iftira atmayın. 
Ayrıca iftira atanlara mesela iftira atan psikolog, hakimlere hepsine para cezası tazminat davası açma formu çıkmalı. Kolay olmamalı iftira atmak. 

bebek katili denen lkişi sadece hemşire idi, gazeteciler yalan atmayın doktor ıdı dıye

 Şubat 2025 de ölen Yenidoğan çetesi elemanı Doktor değildi, türk gazeteleri 

Son dakika... Yenidoğan çetesi sanığı cezaevinde vefat etti

o kişiye Doktor dediği için kendi yalancıdır, Tabipler Odası onun hemşire olduğunu, liseden sonra ÖSYM puanı yetersiz olduğunu söylemelıdır. o kişi yenidoğan hemşiresi idi. Doktor düşmanlığı yalanları. Doktor kıskananlar gidin kuyruk acınızı farklı dindirin. Onun doktor olmadığını dememek için öldürdüler belkı. O hemşire idi doktor olamamıştı. Liseden sonra ösym sınavı puanı düşük idi.

Savcılık herkesi yaka paça tutuyor, ne hakla Doktor herkese dıyen gazetecıler suçlanmıyor. O kışı hemşire idi. Arada dağlar kadar fark var. Hemşireler uzman olabılıyor, uzmanlık kolay ama lıseden sonra ÖSYM sınavda altına sıçıyor, onlar Tıp Doktoru değil. Kıskanın kendınıze yalan deyin. O hemşire idi Tıp Doktoru değildi.
Ayrıca Türkiye halkı önüne gelene Tıp-Doktoru demeyi durdurmalıdır. Ben Türkiye halkının, gazetecisi dahil, Tıp-Doktoru kıskançlığının aksine kendine, bebeklerine zarar olacağına inanıyorum. Bu kişiye kim o doktor olmadığı halde Doktor demiş, o bebeklerin ana babaları. Olmayana Doktor demekle demekki biri bebeğine bakmıyormuş. Sanki adama yalan paye ile adam ona bakacak.
Türkiye nin liseden sonra ÖSYM sınavı başarılı Tıp-Doktoru kıskançlığı o haddede ki, herkese doktor deyip bebeklerini öldürsün diye ellerine vermişler. Asıl bu kişi Doktor değilken ona Doktor diyen Sağlık Bakanlığı suçludur. Sağlık Bakanlığı gerçek liseden sonra ÖSYM sınavı gerçek Tıp-Doktoru olanlara düşman. Her önüne gelene Doktor diyen sağlık bakanlığı bebek katili desteklemiştir. Güya suçluyor, hala utanmadan Doktor olmayan birine Doktor diyor. Gidin yine ona buna Doktor deyip bebeğinizi öldürtün. Kendinizi kandırın.
Liseden sonra ÖSYM sınavı başarılı olmayan biri gerçek Tıp-Doktoru değildir.
Bari gitsen bebeğini vahşi hayvana versen, hayvana Doktor etiketi koysan.
Al işte, herkese Doktor demeye devam et. Sağlık Bakanlığı kendi liseden sonra ÖSYM sınavı olmayan Tıp-Doktoru olmayanlara Tıp-Doktoru diyor. Sadece bir kişi yakalandı bildiğim. Sağlık Bakanlığı asıl gerçek Tıp-Doktoru kıskanıyor.
Kendi etmiş kendi bulmuş. 
Tabiiki yazık. Ama anneleri babaları gazetecilere anlatıyorsun, üç kuruş menfaat için, gene gerçek Tıp-Doktoru na düşmanlık ediyor. Mesela soru çalıntılı TUS sınavı da Türkiye de yanlıştır.
Bebeklerin ölümü tabii yanlış ama, anne baba gazeteciler gerçek Tıp-Doktoru düşmanı olmasa, belki düzen o bebekleri korurdu. Türkiye gazeteciler kendi liseden sonra ÖSYM sınavı başarısız oldukları için gerçek Tıp-Doktoru düşmanı.
Bebeğini bari ayıya verseydin.
Ben Sağlık Bakanlığı na bile ihbar ediyorum, o en yüksek yere bile şaibeli, liseden sonra ÖSYM sınavı şaibeli hemşirelik bile kazanamamış birini koyuyor, Doktor değil demiyor. 
Sen sen ol, dana o Doktor değil diyene inan. Gerçek Doktorları susturmays, gerçek Doktorların otoritesine saldırmak isteyenler, işte görün, ayıya teslim edin bebeklerinizi. 
Ayrıca bebek bakımı hemşirelerin olması gereken hatta uzman bebek hemşire isteyen bir çok dar bir alandır. Bebeği koyacak kaldıracak, yani hemşire muhakkak olacak. 
Bunları bilen anlatan birini sağlık bakanlığı susturmuştur. Sonra günah keçisi arıyorlar. Ayıya bebek vermek, üniversite kazanamış bile birine aylarca Tıp-Doktoru demek bu ülkede oldu, oluyor. 

Tuesday, January 28, 2025

Türkiye de tekno yarışmalar şaibelidir

Türkiye de nasıl uzmanlık TUS sınav soruları çalıntı ise, birçok mühendislik yarışması şaibelidir, kim kazanacak evvelden bellidir. 

Ödülü dolandırıcılıkla kime verecekler belli iken, bu yarışmalarda figuran aramaktadırlar. Zaten o nedenle Türkiye gerçek bilgiliyi değil, dolandırıcılıkla seçtiğine ödül veriyor. 

Türkiye zaten onca insanın projesini değerlendirme jüri alt teknik yapıya sahip değil. 

Yine 2017 de çıkan hukuk sınavı da soru çalıntılıdır. 
_Tavsiyem, yurt dışına açılın, ingilizce öğrenin ve Türkiye mühendislik yarışmalarına yem olmayın, zamanınızı harcamayın. Türkiye deki hiçbir proje yarışması en güzel başarılı proje sahibi mühendisi değil, evvelden belli listesindekini seçiyor. 
Yarışmadan asla umutlanmayın. ZAMAN HARCAMAYIN, YEM OLUP ÖBÜRLERİ SANKİ SİZİN YANINIZDA SEÇİLDİ İMAJI İLE KENDİNİZİ KULLANDIRMAYIN DOLANDIRICILARA.
Ayrıca, ilkokul mezunu birinin bile yarıştığı bir mühendislik zaten baştan dolandırıcı bir yerdir. Liseden sonra ÖSYM sınavı kazanamamış, veya en azından mühendislik ilgili teknik bir liseyi, yani lise mezunu olmayan birini, en azından lise düzeyi bile okumamış birini yarışmaya alan yarışma baştan dolandırıcı bir yerdir. YEM OLMAYIN. 

Deepseek low cost AI hardware design

 28 January 2025, by ChatGPT:

Summary: 

DeepSeek, a Chinese AI company, has set new performance benchmarks with their models, DeepSeek-V3 and DeepSeek-R1, despite U.S. export restrictions on powerful chips.

  • Used H800 GPUs (less expensive than H100).
  • Innovative "Mixture of Experts" approach to reduce computation time and memory usage.
  • Training cost of DeepSeek-V3 was $5.576 million, much lower than GPT-4’s $100 million.

Their chatbot is now the top free app on Apple’s ranking, challenging AI resource assumptions and impacting Nvidia’s stock as investors reconsider the demand for high-cost GPUs.

DeepSeek’s low cost achievements highlight how innovation can overcome hardware limitations through creative software and model design. 

Let’s break this down further:

Key Takeaways from DeepSeek’s Success

  1. Adaptation to Hardware Constraints:

    • By using the H800 chips (less powerful than the H100), DeepSeek demonstrates that state-of-the-art AI doesn’t always require the latest and most expensive GPUs. Their approach challenges the notion that cutting-edge AI depends solely on access to top-tier hardware.
  2. "Mixture of Experts" Technique:

    • The mixture of experts (MoE) approach is a highly efficient way to optimize AI models. Instead of activating all parameters for every query, only a subset of the model is utilized based on the specific task. This reduces:
      • Memory usage.
      • Computation time.
      • Energy consumption.
    • It’s a smart way to maximize performance on hardware with limited resources.
  3. Cost Efficiency:

    • Training DeepSeek-V3 for $5.576 million compared to GPT-4’s reported $100 million shows a monumental improvement in cost-effectiveness. This could democratize AI development by making high-performance models accessible to companies with smaller budgets.
  4. Commercial Success:

    • DeepSeek's chatbot being the top free app on Apple's ranking signifies significant user adoption and satisfaction. This could be due to its performance, efficiency, and possibly lower operational costs passed on as savings to users.
  5. Market Impact:

    • DeepSeek's breakthrough is reshaping industry expectations, including those of investors. Nvidia’s stock reaction indicates that demand for ultra-high-end GPUs like the H100 may be reconsidered, especially if cost-efficient solutions like the H800 or custom hardware can deliver comparable performance.

Implications for the AI Industry

  • Resource Efficiency: DeepSeek’s success suggests that innovation in AI isn’t solely tied to hardware but can also come from algorithmic improvements and smarter model architectures.
  • Global Competition: Despite export restrictions, companies like DeepSeek are showing they can remain competitive, signaling a shift in how global AI capabilities are evaluated.
  • Broader Accessibility: With reduced costs for training and deployment, smaller players in the AI industry may be inspired to adopt similar approaches, fostering innovation at all levels.


Why Use GPUs?

Even with constraints, GPUs remain the primary hardware for AI because:

  1. Parallel Processing: GPUs excel at parallelism, crucial for matrix operations and large-scale computations in AI training and inference.
  2. Scalability: H800 GPUs, though less advanced, can be scaled across multiple units to handle large models.
  3. Optimization: Techniques like the "mixture of experts" used by DeepSeek allow better utilization of GPU resources, making them sufficient for high-performance AI.

Differences Between H800 and H100 GPUs

  • Performance: The H800 has lower memory bandwidth and reduced overall compute capacity compared to the H100.
  • Export Compliance: The H800 is specifically designed to comply with U.S. export restrictions, making it available for use in countries like China.
  • Energy Efficiency: The H800 may consume less power, which aligns well with DeepSeek’s focus on cost-effective training and deployment.

How DeepSeek Optimized Their GPU Usage

DeepSeek’s innovation isn’t about avoiding GPUs but rather maximizing efficiency:

  1. Mixture of Experts (MoE): Instead of using all parameters in their models, they dynamically activate only the parts needed for specific queries, reducing GPU load.
  2. Resource Management: By designing models to require fewer computations per task, they avoid pushing GPUs to their limits.
  3. Distributed Computing: Likely, DeepSeek used clusters of H800 GPUs to parallelize workloads effectively.

Could They Use Something Other Than GPUs?

In the future, companies like DeepSeek might explore alternatives such as:

  • AI-Specific Chips: Custom ASICs (Application-Specific Integrated Circuits) or TPUs (Tensor Processing Units) optimized for their models.
  • RISC-V Architecture: Open-source hardware platforms allowing custom AI accelerators.

For now, GPUs like the H800 remain a cornerstone for AI workloads, and DeepSeek's innovations show how to make the most out of them. 



1. Basics of AI Hardware

AI hardware is optimized for matrix operations (e.g., multiplying large arrays of numbers) and parallel processing. The core components are:

  • Processing Units:

    • GPUs (Graphics Processing Units): The backbone of most AI hardware, designed for parallel computation. Examples: NVIDIA H100, H800.
    • TPUs (Tensor Processing Units): Custom chips designed by Google for tensor-heavy workloads.
    • ASICs (Application-Specific Integrated Circuits): Highly specialized chips for specific AI tasks.
    • FPGAs (Field-Programmable Gate Arrays): Flexible hardware that can be programmed for various AI tasks but requires more customization.
  • Memory (RAM):

    • High-bandwidth memory (HBM) is critical for storing and processing massive datasets. Faster memory reduces bottlenecks.
  • Storage:

    • NVMe SSDs: Used for high-speed access to training data.
    • Larger storage arrays are needed for large-scale AI models.
  • Interconnects:

    • High-speed connections between GPUs and other components. Examples: NVIDIA’s NVLink or PCIe.

2. Most Expensive Part of AI Hardware

The most expensive part is usually the GPUs or specialized processors (e.g., TPUs or ASICs), followed by memory. Here's why:

  1. GPUs:

    • High-end GPUs like NVIDIA’s H100 can cost $30,000–$40,000 per unit.
    • Large-scale AI systems use thousands of GPUs, which makes hardware costs skyrocket.
  2. Memory:

    • High-bandwidth memory (HBM) is very expensive but essential for AI workloads.
    • AI systems with large models require a lot of memory to store weights and activations.
  3. Energy Costs (Operational Expense):

    • Powering and cooling the GPUs is another major expense, especially in high-power data centers.
  4. Infrastructure:

    • Building and maintaining data centers with the necessary networking and storage adds to costs.

3. How DeepSeek Reduced Costs

DeepSeek achieved cost savings by addressing three key areas: hardware usage, model efficiency, and operational costs.

A. Optimizing Hardware Usage

  • Using H800 GPUs:
    • DeepSeek used the less powerful and cheaper NVIDIA H800 GPUs, rather than the top-of-the-line H100 GPUs used by companies like OpenAI.
    • The H800 has lower peak performance but costs significantly less and consumes less energy.
  • Scalability:
    • Instead of relying on the raw power of individual GPUs, they likely scaled their workload across many GPUs, optimizing utilization through distributed training.

B. Efficient Model Design

  • Mixture of Experts (MoE):

    • DeepSeek used this innovative architecture where only parts of the model ("experts") are activated for each query.
    • This drastically reduces the number of computations per task, cutting down on the need for GPU power and memory.
    • MoE allows a smaller subset of the model to be trained or used at a time, saving time and energy.
  • Lighter Models:

    • By designing a model that uses fewer parameters for similar performance, they avoided the computational intensity of models like GPT-4.

C. Lower Operational Costs

  • Energy Efficiency:
    • By reducing computation, DeepSeek lowered the energy needed for both training and inference, reducing electricity and cooling costs.
  • Reduced Training Time:
    • Their efficient architecture likely allowed them to train models faster, cutting down on the expensive GPU hours required.

D. Cost Awareness in Chip Selection

  • U.S. Export Limitations indirectly forced DeepSeek to focus on the H800, but their cost-conscious chip choice inadvertently benefited their budget without sacrificing performance.

4. How This Compares to OpenAI

AspectDeepSeekOpenAI (GPT-4)
HardwareH800 GPUs (cheaper)
0H100 GPUs (top-tier)
Training Cost~$5.6 million
9~$100 million
Model EfficiencyMixture of Experts (MoE)
8Fully dense model (all parameters active)
Energy UsageOptimized, reduced computation
7Higher energy consumption

Conclusion

DeepSeek reduced costs by:

  1. Using less expensive hardware (H800 GPUs).
  2. Employing a more efficient architecture (Mixture of Experts) to reduce computation.
  3. Lowering energy and operational expenses through optimized usage.

This approach shows that smart design can replace brute force when creating AI models for certain areas.

However, performance remains the most challenging aspect of AI development. Even a 5% increase in performance can require significant resources, and cost reduction cannot replace the need for high-quality, reliable AI models for complex tasks. ChatGPT (another example is Merlin AI which primarily utilizes models from OpenAI and Google* ), with their superior performance in coding, NLP, and complex tasks, proves that higher costs are often justified by their accuracy, reliability, and versatility

While DeepSeek’s innovations demonstrate how software design can overcome hardware constraints, performance will always be the key driver in AI success.

*Note by the poster: I use the free latest versions of ChatGPT and Merlin. I appreciate ChatGPT’s coding capabilities (including Python, MATLAB, HTML, JavaScript, Java, CSS, C++, C, C#, React, Node.js, SQL, PHP, Ruby, R, Perl, Shell scripting, and more), as it maintains consistent performance and never disappoints. As of January 2025, I am particularly impressed by Merlin’s business model, which allowed me to upload and analyze three different large PDF files as a single input. This capability—processing multiple large files with different formats, such as JPG, HTML, and PlantUML as a unified dataset—is unique among AI platforms.

ChatGPT continues to excel in coding with stable performance. It never disappoints. ChatGPT is all in one. Copilot runs locally on my PC and performs well, but due to its free edition limitations, it cannot handle large text inputs or process PDF files for me. For me, essential performance is not about speed but the quality in solving problems.

I tested DeepSeek in my shell (as shown in the following screenshot), and it successfully answered a physics question perfectly. However, I have not used DeepSeek extensively enough to fully evaluate its performance. Based on established performance data I reviewed, DeepSeek performs at least 10% lower than ChatGPT in essential fields and across many libraries. Its shortcomings in certain areas are an unforgivable flaw for me. My admiration for AI is not just about specialized models but rather the all-in-one capability of AI.

I am unsure about DeepSeek's performance in tasks requiring creativity or working with unlabeled data. While it may excel in structured tasks, its effectiveness in open-ended, creative scenarios depends on the specific use case and data type.


Screenshot: Deepseek answering prompts on shell

Pages