@elonmusk@elonmusk curious about something which you may or may not have explored - water as a data storage medium? There is something special about water so it’s not impossible that if the source of life is water it could not be used to store code
@ArtemXTech You can actually ask Claude to store memory file before you close the session; kill the terminal restart again and say, Claude retrieve memory and it does - works for me
On balance I feel codex is fast but at a cost - in most instances codex doesn’t read the skills docs , ignores commit hooks and most of the time over complicates things; if you are building a simple app then I can see the benefits; however if you are approaching enterprise saas level- I prefer the slow meticulous approach of Claude- although I have noticed that the new model of opus is sometimes given to shortcutting things instead of solving the hard problems (if you are not careful you end up with a lot of stubs) - but the most painful part is Claude despite compacting still ends up with context size limit errors or Bun runtime crash
@ahmednasirlaw You have been to Addis, most of your readers have not; do you honestly think any governor, let alone Sakaja, can bring Nairobis transformation to Addis level? Perhaps this was a clear recognition that sometimes change requires bold thinking
@OpenAIDevs Codex it’s fast .. until I asked it, have you read the Md files? Then it was like ahem no! For complex topics that require step by step approach, eg enterprise saas I am
All for slow and steady I.e Claude; but for small apps, I can see why one shorting a feature app would be ok
I don’t want to be alarmist but something strange 2day, not with clawdbot but with Claude code on my terminal; on Friday I finished my session, and run an upgrade and left with no new sessions opened. This morning I found it was up for 20hrs and it refused all my commands to stop. I had to kill the terminal and restart.. but I have no clue what it was doing those 20hrs whilst it was active.
@burkov I was able to solve this by keeping the files smaller e.g if you have monolithic code it will struggle, however if your tsx or ts files below 1200 lines it can read them in blocks of 400lines which i think is within the context limit of 2500 or so tokens that seem to trigger grep
@TosinOlugbenga Tosin- I am building enterprise Saas for asset management, so not your typical technical person for your role, but I think we can collaborate to buid this for African markets - asset managers, insurance, pension plan and banks. I am based in Luxembourg
@Dominus_Kelvin@cyberseb_@AfroIndieHack Hey guys I am based in Luxembourg and I am keen to build an Africa focused product company. Should we link up on a teams call at some point? I am on linked in - https://t.co/EXUkiL6Qgw
Satya Nadela is basically describing the death of the traditional SaaS model.
Explains the AI agentic future, and where the "value" lives.
Because business logic is moving from the software application to the AI agents.
Currently, you buy software for its specific features and rules.
Nadella argues that in the future, software apps will essentially become dumb databases ("CRUD") or simple tools.
The AI Agent will hold all the intelligence, orchestration, and reasoning, simply updating the databases as needed. The software becomes a commodity; the AI becomes the "brain" and the worker.
-----
Video from Bg2 Pod Youtube Channel (link in comment)
@levie I would argue the opposite; you assume software as a commoditised tool but the reality is with agentic systems software becomes personal - customized; software changes from user input fields, tabs etc to data stream ingestion & custom dashboards within an agentic architecture
@bprintco What you articulated there is what i call intent expression with context aware interface: you see today everybody vibecoding is making a copy of what exists instead of rethinking the interface - forms and tabs were designed because humans needed to input fields