Ajanaku. 🇹🇼 .

13.2K posts

Ajanaku. 🇹🇼 . banner
Ajanaku. 🇹🇼 .

Ajanaku. 🇹🇼 .

@4amAJ

no Givenchy I’m a G

Katılım Ağustos 2018
473 Takip Edilen476 Takipçiler
Ajanaku. 🇹🇼 .
@MsAdaO The leap to AI is nothing like any other tool before it, and she’s not wrong, AI can do any knowledge work, that’s not just being a tool, the problem then being quality of output and these are still the early days
English
0
0
3
1.5K
Beyoncé follows her
Gosh this take 🤦🏾‍♀️ Upskilling has always meant adapting to new tools. People said the same thing about spreadsheets, the internet and automation. Refusing to use AI because you may lose your job is like refusing to learn email in 1998 because secretaries used to print memos
Crochet Creator🧶@OreAkinde

This is scary to me and not in a good way, because you’re basically training the ai to take your job and eventually they’ll downsize. That’s not upskilling. Lastly, i am worried that so many people can say the same, soonest companies lose the human input and be relying on bots for everything 🫩

English
10
256
804
122K
L
L@lanreadelowo·
Bun’s rust rewrite has 10K unsafe calls 💀💀💀💀
English
8
0
26
5.2K
Uncle Bob Martin
Uncle Bob Martin@unclebobmartin·
@shashank__says Keeping the structure scalable is part of the engineering challenge. It always has been.
English
2
0
32
1.8K
Uncle Bob Martin
Uncle Bob Martin@unclebobmartin·
Authoring code by hand HAS GONE AWAY. Engineering module structure and architecture has not.
English
135
134
1.3K
86.5K
Joel 🇦🇺
Joel 🇦🇺@ptr_to_joel·
if mfers aren't gonna read the code their llm produces then why not just have them write it in assembly since they're better than humans right
English
9
4
92
9.9K
Uncle OJ
Uncle OJ@uncooloj·
Haha yes yes!! I solve this by having them review each others work. When I’m done with Codex, I ensure I have Claude review it and ensure it knows Codex did the initial task (might be redundant but I do it still). This saves me a lot of these tiny reviews and poses them to deeply review and rethink each others work before I step in. @coder_blvck’s skill has made it easier for them (Codex and Claude for me) to just look at the entire conversation thread of each other as against me having to copy context from one to the other.
English
3
1
1
994
¿kofo?
¿kofo?@kofookesola·
Re: LLMs, you see those tiny hallucinations they make on the lowest level when working on a large task? Those break everything. Today alone i have had to read through the thinking process for agents and correct some decision they made, wondering what would have happened if i didn't
English
5
6
38
6.4K
Kristoph
Kristoph@kristoph·
@edandersen He is only really describing what he built, not the company as a whole - quite a few of the pieces are open source. And honestly his implementation is fairly simple / straightforward ( and a bit naive ) so I don't think any special secrets are being exposed.
English
4
0
316
44.8K
Ed Andersen
Ed Andersen@edandersen·
Incredible video by randomly sacked Atlassian engineer telling all about the entire company Love this genre, like LinkedIn green banner with zero fcks given
Ed Andersen tweet media
English
88
842
18.1K
3.2M
Ajanaku. 🇹🇼 .
I’ve been wondering what really makes services more scalable than monoliths Except you have jobs and stuff in that application “unused code sitting in memory isn’t necessarily expensive”
English
0
0
0
35
Ajanaku. 🇹🇼 .
@bvchidra That’s why you have a data layer in your application, except there’s possibility of tampering from external sources There’s also the fact that if those requirements change, extending the implementation in your db will not be easy or safe
English
0
0
0
10
bvchidra
bvchidra@bvchidra·
im reading takes on here and u can see most people are tweeting directly from chatgpt. some business logic are risky to put only in code u have to enforce it, not only in code, not only at the infra level, both. anyone who says otherwise doesn't know what they are saying
English
3
4
14
766
Johnmark Obiefuna
Johnmark Obiefuna@jayhemz·
Telecoms and insurance companies with complex policy rules, banks and financial institutions, looking at this tweet and laughing. It's clear that a lot of dev experience isn't holistic enough.
os@segun_os_

too much idea want to finish you people on this twitter. some people are writing their business logic in their dBs 😂 how do you handle horizontal scaling, sharding and replication? how do you ensure a highly available system if your business logic is tightly coupled-

English
12
17
108
16.7K
Ajanaku. 🇹🇼 .
@PrebsTamar @segun_os_ And it’s funny how we’re all ignoring the elephant that most systems need to communicate with non-owned services over a network, how are you doing that from your db?
English
0
0
0
13
¿kofo?
¿kofo?@kofookesola·
We killed the best thing to happen to the computing age since the open internet (which we have also killed), source code sharing.
English
1
0
5
819
¿kofo?
¿kofo?@kofookesola·
GitHub actions runners are open source. This is another proof that open source is going to slowly die especially for infrastructure providers. Ai has brought the barrier of searching source code for exploits to zero, and if that is the case, any software with its source shared can be compromised (unless you’re some god than can protect their code from every exploit).
International Cyber Digest@IntCyberDigest

🚨 How the TanStack npm attack actually happened: 1. Attacker opened a normal-looking pull request (#7378) on the TanStack repo. 2. GitHub automatically ran CI tests on that PR. 3. Code inside the PR stole the workflow's GitHub Actions Cache write token during the test run. 4. The attacker used that token to plant poisoned files in the shared build cache. The PR could be closed afterwards. The poisoned cache stays. 5. The official release workflow later pulled from the cache, baked the malicious files into the build, and signed and published 84 malicious package versions to npm.

English
11
7
37
8.9K
Ajanaku. 🇹🇼 .
@bahdcoder These are mostly process skills, full-stack in 1 year sounds somehow but this isn’t the angle
English
0
0
1
188
Frantz Kati
Frantz Kati@bahdcoder·
Sure. same problem with most people getting in the industry. Learn all the skills you want. You're still useless if you do not have a deep understanding of topics like: - Version control & team collaboration - Scale, system design, system architecture - Product engineering - Engineering cycles - Customer feedback to production software - Feature flagging & testing in production - Leadership, feedback giving and receiving - CI & CD pipelines, PR reviews - Analytical thinking, product management - Product strategy, user experience, communication And this does not even scratch the surface. And no. You can't learn this in 1 year.
English
6
8
50
2.9K
Greg Mushen
Greg Mushen@gregmushen·
@TheStalwart @nic_carter It's always been a tough problem. The difference today is that their bosses can use AI to inspect each contribution. That process can read the code and all of the changes and help make a determination. So raw token count itself will only last a short time.
English
1
0
5
1K
Joe Weisenthal
Joe Weisenthal@TheStalwart·
The FT says that Amazon employees are doing random unnecessary task automations to consume tokens and to show their bosses that they're using AI more ft.com/content/8ee0d3…
English
236
652
6.7K
2.6M