GROK 2 vs. LLAMA 3.1 - Cloud vs Home Server Ai Testing

GROK 2 is a surprisingly good beta mini version model that X released this week. I run a battery of tests with xAI's latest, Grok 2.0, against Meta's cutting-edge Llama 3.1 70B that is running locally. We do a variety of user centric tasking tests as well as some coding.

Be sure to 👍✅Subscribe✅👍 for more content like this!

Please share this video to help spread the word and drop a comment below with your thoughts or questions. Thanks for watching!

Chapters
0:00 Meta LLAMA 3.1 vs xAI GROK 2
1:45 Setup VS Code and Python for Ai
4:20 GROK 2 Python Game Creation
6:37 LLAMA 3.1 Python Game Creation
8:17 GROK 2 Creative Storytelling
10:42 LLAMA 3.1 Creative Storytelling
12:02 GROK 2 Deductive Reasoning
13:10 LLAMA 3.1 Deductive Reasoning
14:26 LLAMA 3.1 Count Words in Sentence
15:22 GROK 2 Count Words in Sentence
16:22 GROK 2 Decimal Precision and Counting
16:35 LLAMA 3.1 Decimal Precision and Counting
17:27 GROK 2 Recipe from Ingredients
18:47 LLAMA 3.1 Recipe from Ingredients
20:55 LLAMA 3.1 Food Budgeting and Planning
23:16 GROK 2 Food Budgeting and Planning
26:06 GROK 2 Fitness Coach and Planner
28:13 LLAMA 3.1 Fitness Coach and Planner
30:43 Home Ai Server Privacy vs Cloud Ai Provider
32:14 Conclusion

*****
As an Amazon Associate I earn from qualifying purchases.

When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
*****

[EDGSV-fm]

🎯 Key points for quick navigation:

00:00:00 *🥅 Testing Overview*
- Introduction of AI models: Grok 2.0 vs. Llama 3.1,
- Emphasis on practical usability in testing,
- Mention of the beta status of Grok 2.0.
00:02:11 *💻 Visual Studio Setup*
- Instructions for installing Visual Studio on Windows,
- Setup guidance for Python projects and PyGame installation.
00:04:18 *🎮 Game Development Test*
- Grok and Llama 3.1 tasked to create a game called "Flappy Block",
- Grok showed better performance in producing functional Python code,
- Initial outputs required debugging from both models.
00:08:29 *📖 Creative Storytelling*
- Models tasked with creating a humorous story,
- Output was verbose and not particularly funny,
- Neither model excelled in succinctness or humorous storytelling.
00:12:49 *🐱 Deductive Reasoning Test*
- Scenario involving a cat and a carrier tested models' deductive reasoning,
- Neither model provided a correct or insightful solution,
- Highlights challenges in AI's deductive abilities.
00:14:42 *🧮 Numerical Reasoning*
- Basic numerical comparison challenge,
- Both Grok and Llama 3.1 correctly identified the larger number.
00:17:19 *🍳 Recipe Creation*
- Models tasked with creating a recipe from limited ingredients,
- Both models successfully generated recipe instructions,
- Llama 3.1 provided additional context with serving size and tips.
00:20:59 *🛒 Budget Meal Planning*
- Creation of a grocery shopping list and weekly meal plan,
- Emphasis on budget and caloric needs,
- Pricing and specifics of the meal plan noted to be inaccurate or outdated.
22:41 *📊 Inaccuracies in Nutritional Guidelines*
- Discussion of caloric intake recommendations for a 40-year-old male.
- Assessment of budget shopping lists and recipe nutritional information.
25:57 *🏋️ Analysis of Fitness Plans*
- Evaluation of exercise instructions and plans for a year-end weight target.
- Critique of calorie burn estimates and effectiveness of workout plans.
31:02 *🔒 Privacy Concerns with AI Models*
- Consideration of privacy implications with cloud-based AI.
- Comparison between Grok 2.0 and Llama 3.1 regarding data privacy.

Made with HARPA AI

[KastanDay]

10:40 - models often struggle with multi-turn conversations. I highly recommend starting a new conversation, especially when changing subjects dramatically (from coding to storytelling).

[dllsmartphone]

Messed up "video position shortcut" in description (Home Ai Server Privacy vs Cloud Ai Provider)

[nufh]

Do you have a guide on how to host Llama 70B like in this video? I couldn't find it in your previous videos.

[mikeygomes]

Nice rig, how did you put it together

[termino]

can you share soucre code ui for LLAMA 3.1

[大支爺]

Run locally? not both
Uncensored? not both
No comparison points.