Experimenting with ChatGPT

Looks like a Facebook Uncle Type of Argument But It's Difficult to Deny that Jewish/Satanic Symbolism is actually a thing.

The Open Ai Icon:

 

A Shia LLM.

 

https://m.rediff.com/news/report/in...-model-under-rs-100hour-vaishnaw/20250131.htm

India too wants to give DeepSeek a run for its money...

 

More importantly Deepseek is an open source model, some models can also be run locally which are on par with GPT O1.

 

https://www.deepseek.com/

i asked deepseek which was better. the reply:

----
The question of whether DeepSeek or OpenAI is "better" depends on the specific use case, requirements, and context in which their technologies are being applied. Both organizations are leaders in AI research and development, but they have different focuses, strengths, and offerings.

DeepSeek
- Focus: DeepSeek is a Chinese company specializing in AI-driven solutions, particularly in areas like natural language processing, computer vision, and AI applications for specific industries.

- Strengths:
- Tailored solutions for specific industries or regional markets (e.g., China).
- May offer more localized or specialized AI models for certain applications.
- Potentially more cost-effective for niche use cases.

- Weaknesses:
- Less globally recognized compared to OpenAI.
- May have fewer resources or less extensive research compared to OpenAI.

OpenAI

- Focus: OpenAI is a globally recognized AI research organization known for developing cutting-edge models like GPT-4, DALL·E, and ChatGPT.

- Strengths:
- State-of-the-art models with broad applicability across industries.
- Strong emphasis on research and innovation.
- Widely adopted and integrated into global applications.
- Extensive documentation, community support, and developer tools.

- Weaknesses:
- Can be expensive for large-scale or enterprise use.
- May not always offer localized or region-specific solutions.

Key Considerations
1. Use Case: If you need a globally recognized, general-purpose AI model, OpenAI might be better. For specialized or localized solutions, DeepSeek could be more suitable.

2. Budget: DeepSeek might offer more cost-effective solutions for specific needs, while OpenAI's advanced models can be more expensive.

3.Localization: DeepSeek may have an edge in understanding and catering to specific regional or cultural contexts, especially in China.

4.Innovation: OpenAI is often at the forefront of AI research, making it a better choice for cutting-edge applications.

Ultimately, the "better" choice depends on your specific needs, such as the type of AI application, budget, and geographic focus. Both organizations have their unique strengths and can excel in different scenarios.

=====================

 

https://www.bbc.com/news/live/cjr85l2e4l4t

We were all just getting accustomed to Trump's round up of a 1000 illegals and a deportations plane being sent back, and suddenly this morning cnn and bbc headlines are all about some shoestring Chinese AI project throwing a monkey wrench into the American one. Very interesting...

 

This is an interesting video. I’m wondering, what if someone takes an open LLM like LLaMA and trains it with at least 20k tokens per day on Islamic poets like Rumi, Busiri, Ala Hazrat, Yusuf Nabahani, etc.? Would the resulting poems capture any essence of their personalities?

 

https://www.bbc.com/news/articles/cd0el3r2nlko

Hahaha... no foul play!

 

Artificial Intelligence or Artificial Wahabiism? someone sent me this auto-generated appraisal on the hadith implying that 9/10'ths of barakah are in trade:

---

The narration “Nine-tenths of sustenance is in trade” (التِّسْعَةُ أَعْشَارُ الرِّزْقِ فِي التِّجَارَةِ) is not found in the widely recognized collections of Hadith such as Sahih Bukhari or Sahih Muslim. It is attributed to less authoritative sources, and its authenticity is considered weak (da’if) or untraceable. Some scholars attribute it to compilations such as those by Imam Suyuti or other minor Hadith collections.

Scholarly Perspective
• The narration has been cited by scholars in discussions on the virtues of trade but is often categorized as weak due to the lack of a strong chain of transmission.
• Despite its weak status, the sentiment aligns with Islamic principles, as trade and lawful earning are strongly encouraged in Islam.

If you are looking for an authentic Hadith on trade, consider the following:

“The truthful and trustworthy merchant will be with the Prophets, the truthful, and the martyrs.”
(Sunan al-Tirmidhi, Hadith 1209; classified as Hasan)

This Hadith emphasizes the spiritual rewards for honesty and trustworthiness in business dealings.

 

I have seen many incorrect taweezat, but this one tops the list. It follows none of the rules, it does not add up in every dila at all.

What could have been done here is making a murabba', that adds up to the sum 786, but completed by adding 7 or 9 instead of whatever was presented there(usually one adds 1 per square etc).

In this case you'd do {786-210} /4 = 144

Then you complete the murabba naqsh by adding 7 following the atishi sequence.

For 9 you'd do (786-270)/4 = 129

And complete by adding 9 every square in the sequence.

 

They were already trained through YouTube videos. Many content creators have already discussed this, and some have even moved to platforms like Rumble and Nebula.

 

lol ... it will be even more fun once these data-hungry bots start trawling public domain YT videos for content to train on (if they have not already) - so many aamils will go out of business!

 

Tawiz Naqsh for Fever:

1. Center: Write the verse from the Quran associated with healing:
   • Surah Al-Isra (17:82):
     "Wa nunazzilu minal Qur’ani ma huwa shifaun wa rahmatun lilmu’mineen."
     (Translation: "And We send down from the Quran that which is a healing and a mercy to those who believe.")
2. Corners: Write the Names of Allah associated with mercy and healing:
   • Ya Shafi (O Healer)
   • Ya Rahman (O Most Merciful)
   • Ya Rahim (O Most Compassionate)
   • Ya Latif (O Subtle and Kind)
3. Outer Circle/Edges: Add the phrase Bismillahir Rahmanir Raheem (In the Name of Allah, the Most Gracious, the Most Merciful) repeatedly for protection.
   
   
4. Numerical Grid (Optional): Include a naqsh using the Abjad values of healing-related verses or Allah's names.

Simple Naqsh Grid:
7 9 7
9 786 9
7 9 7

• 786 in the center symbolizes Bismillahir Rahmanir Raheem.
• 7 and 9 are symbolic numbers used in spiritual healing.

---
[this is a screenshot of the naqsh grid presented- AQ]



---

will be a mess once some youtube and tv aamils learn of this chatGPT generated tawiz's!

 

ChatGPT says:
----
Here is the complete list of Allah's 99 Names (Asma'ul Husna) along with their corresponding Abjad values. The values are calculated based on the Abjad system used in Islamic numerology.

# Name Arabic Abjad Value
1 Ar-Rahman الرَّحْمَٰن 329
2 Ar-Raheem الرَّحِيم 289
3 Al-Malik الْمَلِك 90
4 Al-Quddus الْقُدُّوس 170
5 As-Salam السَّلَام 131
6 Al-Mu'min الْمُؤْمِن 136
7 Al-Muhaymin الْمُهَيْمِن 145
8 Al-Azeez الْعَزِيز 94
9 Al-Jabbar الْجَبَّار 206
10 Al-Mutakabbir الْمُتَكَبِّر 662
11 Al-Khaliq الْخَالِق 731
12 Al-Bari' الْبَارِئ 213
13 Al-Musawwir الْمُصَوِّر 336
14 Al-Ghaffar الْغَفَّار 1286
15 Al-Qahhar الْقَهَّار 306
16 Al-Wahhab الْوَهَّاب 14
17 Ar-Razzaq الرَّزَّاق 308
18 Al-Fattah الْفَتَّاح 489
19 Al-Aleem الْعَلِيم 150
20 Al-Qabid الْقَابِض 903
21 Al-Basit الْبَاسِط 72
22 Al-Khafid الْخَافِض 1481
23 Ar-Rafi' الرَّافِع 351
24 Al-Mu'izz الْمُعِزُّ 117
25 Al-Muzil الْمُذِلُّ 770
26 As-Sami' السَّمِيع 180
27 Al-Basir الْبَصِير 302
28 Al-Hakam الْحَكَم 68
29 Al-Adl الْعَدْل 104
30 Al-Latif اللَّطِيف 129
31 Al-Khabir الْخَبِير 812
32 Al-Halim الْحَلِيم 150
33 Al-Azeem الْعَظِيم 1020
34 Al-Ghafoor الْغَفُور 1286
35 Ash-Shakoor الشَّكُور 526
36 Al-Aliyy الْعَلِيّ 110
37 Al-Kabir الْكَبِير 232
38 Al-Hafiz الْحَفِيظ 998
39 Al-Muqeet الْمُقِيت 550
40 Al-Hasib الْحَسِيب 80
41 Al-Jaleel الْجَلِيل 73
42 Al-Karim الْكَرِيم 270
43 Ar-Raqib الرَّقِيب 312
44 Al-Mujeeb الْمُجِيب 55
45 Al-Wasi' الْوَاسِع 137
46 Al-Hakeem الْحَكِيم 78
47 Al-Wadud الْوَدُود 20
48 Al-Majeed الْمَجِيد 57
49 Al-Ba'ith الْبَاعِث 573
50 Ash-Shaheed الشَّهِيد 319
51 Al-Haqq الْحَقُّ 108
52 Al-Wakeel الْوَكِيل 66
53 Al-Qawiyy الْقَوِيّ 116
54 Al-Matin الْمَتِين 500
55 Al-Waliyy الْوَلِيّ 46
56 Al-Hamid الْحَمِيد 100
57 Al-Muhsi الْمُحْصِي 148
58 Al-Mubdi' الْمُبْدِئ 57
59 Al-Mu'id الْمُعِيد 124
60 Al-Muhyi الْمُحْيِي 68
61 Al-Mumeet الْمُمِيت 490
62 Al-Hayy الْحَيُّ 18
63 Al-Qayyum الْقَيُّوم 156
64 Al-Wajid الْوَاجِد 14
65 Al-Majid الْمَاجِد 48
66 Al-Wahid الْوَاحِد 19
67 Al-Ahad الْأَحَد 13
68 As-Samad الصَّمَد 134
69 Al-Qadir الْقَادِر 305
70 Al-Muqtadir الْمُقْتَدِر 744
71 Al-Muqaddim الْمُقَدِّم 184
72 Al-Mu'akhkhir الْمُؤَخِّر 801
73 Al-Awwal الْأَوَّل 37
74 Al-Akhir الْآخِر 801
75 Az-Zahir الظَّاهِر 1106
76 Al-Batin الْبَاطِن 62
77 Al-Wali الْوَالِي 47
78 Al-Muta'ali الْمُتَعَالِي 551
79 Al-Barr الْبَرُّ 202
80 At-Tawwab التَّوَّاب 409
81 Al-Muntaqim الْمُنْتَقِم 630
82 Al-Afuww الْعَفُوُّ 156
83 Ar-Ra'uf الرَّؤُوف 287
84 Malik-ul-Mulk مَالِكُ الْمُلْك 1065
85 Dhul-Jalali wal-Ikram ذُو الْجَلَالِ وَالْإِكْرَام 1001
86 Al-Muqsit الْمُقْسِط 209
87 Al-Jami' الْجَامِع 114
88 Al-Ghaniyy الْغَنِيُّ 1060
89 Al-Mughni الْمُغْنِي 1100
90 Al-Mani' الْمَانِع 161
91 Ad-Darr الضَّارَّ 1001
92 An-Nafi' النَّافِع 201
93 An-Nur النُّور 256
94 Al-Hadi الْهَادِي 20
95 Al-Badi' الْبَدِيع 86
96 Al-Baqi الْبَاقِي 113
97 Al-Warith الْوَارِث 707
98 Ar-Rashid الرَّشِيد 514
99 As-Saboor الصَّبُور 298
Let me know if you need any further clarification or analysis!

---

The Abjad value for Allah (اللَّه) in Arabic can be calculated as follows:

Breaking it down:

• ا (Alif) = 1
• ل (Lam) = 30
• ل (Lam) = 30
• ه (Ha) = 5

Total:
1 + 30 + 30 + 5 = 66

The Abjad value for Allah (اللَّه) is 66.

---

To calculate the Abjad value for "Bismillahir Rahmanir Raheem" (بِسْمِ اللَّهِ الرَّحْمَٰنِ الرَّحِيمِ) in Arabic, we assign numerical values to each letter using the Abjad system.

Breaking it down:
For بِسْمِ (Bism):

• ب = 2
• س = 60
• م = 40
  Total for Bism = 2 + 60 + 40 = 102

For اللَّهِ (Allah):

• ا = 1
• ل = 30
• ل = 30
• ه = 5
  Total for Allah = 1 + 30 + 30 + 5 = 66

For الرَّحْمَٰنِ (Ar-Rahman):

• ا = 1
• ل = 30
• ر = 200
• ح = 8
• م = 40
• ن = 50
  Total for Ar-Rahman = 1 + 30 + 200 + 8 + 40 + 50 = 329

For الرَّحِيمِ (Ar-Raheem):

• ا = 1
• ل = 30
• ر = 200
• ح = 8
• ي = 10
• م = 40
  Total for Ar-Raheem = 1 + 30 + 200 + 8 + 10 + 40 = 289

Final Total:
Bismillahir Rahmanir Raheem = 102 (Bism) + 66 (Allah) + 329 (Ar-Rahman) + 289 (Ar-Raheem) = 786

The Abjad value for "Bismillahir Rahmanir Raheem" (بِسْمِ اللَّهِ الرَّحْمَٰنِ الرَّحِيمِ) is 786, which is a well-known number associated with this phrase in Islamic tradition.

---

The Abjad system assigns numerical values to Arabic letters and is often used in Islamic numerology to explore the mystical or symbolic meanings of words and names.

To calculate the Abjad value for "Muhammad Ali" (محمد علي) in Arabic:

Breaking it down:
For محمد (Muhammad):

• م = 40
• ح = 8
• م = 40
• د = 4
  Total for Muhammad = 40 + 8 + 40 + 4 = 92

For علي (Ali):

• ع = 70
• ل = 30
• ي = 10
  Total for Ali = 70 + 30 + 10 = 110

Final Total:
Muhammad Ali = 92 (Muhammad) + 110 (Ali) = 202

The Abjad value for "Muhammad Ali" (محمد علي) is 202.

This number may hold spiritual or symbolic significance in certain contexts, depending on the practice or interpretation of numerology in Islamic traditions.

 

i still haven't made my zarda. the last time i ever cooked anything was in 2004!

 

Experimenting with ChatGPT is a great way to explore the capabilities of AI. You can try asking it questions, requesting different types of content, or even playing games. What kind of experiments have you been doing with ChatGPT?

 

I understand that Open AI's Strawberry (released earlier this month) is capable of intermediate level reasoning. Reviewers have been buzzing.

https://www.vox.com/future-perfect/371827/openai-chatgpt-artificial-intelligence-ai-risk-strawberry
https://www.tomsguide.com/ai/chatgp...rry-model-from-openai-5-prompts-to-try-it-out

----

Then there's Google's DeepMind which used AlphaGeometry 2 (combined with AlphaProof/LeanCode) to solve 4 International Mathematical Olympiad (IMO) problems, that would have secured it silver medal. It, however, took 3 days to solve the problems, when the contestants get only 2x 4.5 hours to solve 6 problems. For the uninitiated, IMO is considered the hardest problem solving competition aimed at senior high school students.

To my mind, that is seriously impressive. It's worth noting that 2 problems that AlphaGeometry failed to crack were combinatorics related, which require significantly higher order of reasoning.

https://deepmind.google/discover/blog/ai-solves-imo-problems-at-silver-medal-level/
https://www.lesswrong.com/posts/TyC...s-silver-medal-standard-solving-international
https://www.theguardian.com/technol...takes-step-closer-to-cracking-top-level-maths
https://www.maginative.com/article/google-deepminds-ai-scores-silver-at-international-math-olympiad/

 

https://www.bbc.com/future/article/20240912-what-riddles-teach-us-about-the-human-mind

It seems AI can only recognize outer content but is void of even elementary reasoning of kids.

 

“Hey, you should only be downloading US spyware!”

Some have suggested that this move arose because Kaspersky has a history of detecting NATO malware. And recently, revealed a new way to detect the Israeli Pegasus spyware.

 

https://www.bbc.com/news/articles/ceqq7663wd2o

Apparently they're (FAANG) lesser evils than the Russian Kaspersky!