dmesg --follow
[ 68157360.000 ] posts.x: Full workshop on turning feature requests into job statements and prioritizing them with the Kano model: |
[ 68157360.000 ] posts.x: When AI agents level the implementation playing field, building the right thing becomes the actual differentiator. That's the thesis @kentcdodds, a… |
[ 68151420.000 ] posts.x: I updated my article on Codex context windows. The old trick works again, on GPT-5.6 this time: set model_context_window in a profile and you get… |
[ 68144100.000 ] posts.x: @GeoffreyHuntley |
[ 68144100.000 ] posts.x: @varun_pant_ |
[ 68141880.000 ] posts.x: Full workshop on the sandbox failure modes E2B built this CTF around, from runaway processes to orphan cleanup to sandbox lifecycle management: |
[ 68141880.000 ] posts.x: Most sandbox bugs don't show up when you're running one of them. They show up once you're running thousands. @badphilosopher, an AI engineer at E2B… |
[ 68138220.000 ] posts.x: This flash memory shortage is no joke! I paid $180 for this 2TB SSD in December 2024. It's now $570. Usually drive prices *decrease* over a similar… |
[ 68133780.000 ] posts.x: Full workshop on building the right thing before building it right, including the mom test questions and Don Norman's take on user error: |
[ 68133780.000 ] posts.x: Implementation used to be the hard part of software engineering. Now agents handle that fast, so the differentiator becomes deciding what's worth… |
[ 68093280.000 ] posts.x: Full workshop on building an LLM judge from scratch, the biases that make it lie, and the Instagram chatbot incident that shows why it matters: |
[ 68093280.000 ] posts.x: An Instagram support bot once let attackers past two-factor authentication with nothing more than a VPN and a polite request, and 20,225 accounts got… |
[ 68085900.000 ] posts.x: Most engineers still babysit one agent at a time, and "Lifestyles of the AI-Native" from @aiDotEngineer on YouTube makes the case that this is… |
[ 68085960.000 ] posts.x: Full talk on moving from babysitting a single agent session to running a fleet of them, voice coding, loops, goals, hooks, and scheduled tasks… |
[ 68157360.000 ] posts.x: Full workshop on turning feature requests into job statements and prioritizing them with the Kano model: |
[ 68157360.000 ] posts.x: When AI agents level the implementation playing field, building the right thing becomes the actual differentiator. That's the thesis @kentcdodds, a… |
[ 68151420.000 ] posts.x: I updated my article on Codex context windows. The old trick works again, on GPT-5.6 this time: set model_context_window in a profile and you get… |
[ 68144100.000 ] posts.x: @GeoffreyHuntley |
[ 68144100.000 ] posts.x: @varun_pant_ |
[ 68141880.000 ] posts.x: Full workshop on the sandbox failure modes E2B built this CTF around, from runaway processes to orphan cleanup to sandbox lifecycle management: |
[ 68141880.000 ] posts.x: Most sandbox bugs don't show up when you're running one of them. They show up once you're running thousands. @badphilosopher, an AI engineer at E2B… |
[ 68138220.000 ] posts.x: This flash memory shortage is no joke! I paid $180 for this 2TB SSD in December 2024. It's now $570. Usually drive prices *decrease* over a similar… |
[ 68133780.000 ] posts.x: Full workshop on building the right thing before building it right, including the mom test questions and Don Norman's take on user error: |
[ 68133780.000 ] posts.x: Implementation used to be the hard part of software engineering. Now agents handle that fast, so the differentiator becomes deciding what's worth… |
[ 68093280.000 ] posts.x: Full workshop on building an LLM judge from scratch, the biases that make it lie, and the Instagram chatbot incident that shows why it matters: |
[ 68093280.000 ] posts.x: An Instagram support bot once let attackers past two-factor authentication with nothing more than a VPN and a polite request, and 20,225 accounts got… |
[ 68085900.000 ] posts.x: Most engineers still babysit one agent at a time, and "Lifestyles of the AI-Native" from @aiDotEngineer on YouTube makes the case that this is… |
[ 68085960.000 ] posts.x: Full talk on moving from babysitting a single agent session to running a fleet of them, voice coding, loops, goals, hooks, and scheduled tasks… |