debdut

and gpt-oss is 10x better atleast!

AI slop

Enshittification and Political BS has already happened to Mistral

Not even coding, in most stuff, Mistral models are as embarrassing as LLaMa, because they’re actually just LLaMa

This is crazy, you guys basically rebuilt Swift (interpreter instead of compiler) and also the frameworks. This is the level of engineering that does the magic of "it just works". I, personally, am very interested in swift, and would love to work for free for you guys.

Feel free to dunk, report them (will fix them this weekend), and request some features.

51.1999971, 41.1999989, 4.6000039, In this 3 numbers there are total 9+9+8=26 digits, if these 3 numbers are chosen randomly, what's the probability that there are 10 9's with 2 block of 4 consecutive 9's, 5 1's, 4 0's?

Let's approach this step-by-step using combinatorics:

1) First, let's consider the total number of possible arrangements: We have 26 digits in total, and the order matters. So, the total number of arrangements is: 10^26 (as each position can be filled by any digit from 0 to 9)

2) Now, let's count the favorable arrangements:

   a) We need:
      - 10 nines (including two blocks of 4 consecutive nines)
      - 5 ones
      - 4 zeros
      - 7 other digits (26 - 10 - 5 - 4 = 7)

   b) Let's start by placing the two blocks of 4 consecutive nines:
      We have 19 positions to place the first block (26 - 4 - 3 = 19, as we need to leave room for the second block)
      Then we have 15 positions for the second block
      So, there are 19 * 15 = 285 ways to place these blocks

   c) We need to place 2 more nines:
      We have 18 positions left, so there are C(18,2) = 153 ways to do this

   d) Now, we need to place 5 ones in the remaining 16 positions:
      This can be done in C(16,5) = 4368 ways

   e) Next, place 4 zeros in the remaining 11 positions:
      This can be done in C(11,4) = 330 ways

   f) Finally, we need to fill the remaining 7 positions with the other digits:
      There are 7! = 5040 ways to arrange these

   g) For these last 7 digits, we can choose any digit except 0, 1, and 9:
      So we have 7^7 = 823543 possibilities for what these digits could be

3) Putting it all together: The number of favorable outcomes is: 285 * 153 * 4368 * 330 * 5040 * 823543 = 2.51654 × 10^17

4) Therefore, the probability is: (2.51654 × 10^17) / (10^26) = 2.51654 × 10^-9

So, the probability is approximately 0.00000000251654 or about 1 in 397,371,070,190.

Well, serverless was always touted for it’s scaling capabilities, isn’t it the irony here?

What! They changed the title. Tells you something

The subtitle is "The move from a distributed microservices architecture to a monolith application helped achieve higher scale, resilience, and reduce costs." And the article itself mentions the 90% cost reduction. So the title seems pretty much in-line with the original intent.