Exploring AI, software and emerging technology. Sharing independent analysis, practical insights and informed opinions. Open to interesting collaborations.
I wonder if today’s issue on @Revolut is caused by AI-generated work or not. But the implication of this might be deeper, since AI generated software’s quality has been significantly lower based on latest incidents. Degraded performance and outages occur almost weekly. I think we should start emphasizing stability more than the blazing-fast delivery of features that users might not even need and that lower the software’s quality.
anaplirotis(.)gr has 17,000+ registered users and is completely free for teachers.
I need to keep it affordable to run, too. I wrote about moving to Railway, the actual hosting costs, and keeping the app simple as it grows.
https://t.co/qu0rfBxzfO
AI was supposed to make us work less
I've never been this tired in my life
most days I have up to 10 agents running at the same time. one fixing a bug, one writing pages, one refactoring something I already forgot I asked for, 3 other I have no idea what they are doing but i suppose i ask it?
and I jump from conversation to conversation all day
compared to 5 years ago, I probably context switch 100x more
by the end of the day my brain is fried. completely toasted. nothing left, not even for the people I love
I call it the fried brain effect
thought it was just me but almost every founder I talk to lately says the exact same thing
we got superpowers, and somehow the pressure only went up. if you can ship 10x more, you feel like you have to ship 10x more. every single day
I don't have a solution
honestly I have no idea where this goes from here
anyone else feeling this?
Everyone is going crazy about vibe coding their next one million dollar idea. Reminds me of Ryan in The Office series. We live in exciting times…!
I am probably going to use https://t.co/7WqadzgxCq to reverse engineering WUPHF😅
I agree on both points. It’s not that simple to copy such huge projects; there are a lot of moving parts. Although, think about how many features Jira has that are not being used by many organizations. Then orgs can focus on copying only the parts that matter to them. Of course, every organization should weigh the cost of maintaining and building such tools (as llms get more expensive) versus using them as SaaS from another organization.
@benjaminakar What about B2B? Many companies already start building internal tools to replace their third party apps like Jira, Figma etc… I think this will cause a lot of problems for most of the saas companies and their employees.
@mattpocockuk Tactical tornados are not team players, although they might be great as freelancers.
Most organizations need team players, to mentor and give opportunities to others.
I always tell to my team and colleagues - “It’s a marathon, not a sprint”.
I mean, if you know what you are doing, both models are great at producing accurate results… I will agree that open ai models tend to produce more verbose answers which seems exhausting sometimes.
Have you tried Sol (Max or Ultra) instead of Astra?
I found Astra’s harness to be more geared towards the graphics side.
Το https://t.co/iUPR7lZAgm βοήθησε 17.000+ εκπαιδευτικούς να βρουν την υπολογισμένη θέση τους, σχολεία και περιοχές, να δουν επιμελημένα δεδομένα προσλήψεων του Υπου��γείου Παιδείας και να εκτιμήσουν πιθανότητες για τις επόμενες φάσεις. 💙 Ευχαριστούμε για την εμπιστοσύνη!
I saw this image on the web today.
AI makes building an app more accessible. It doesn’t make prod problems disappear.
config hell, permissions, database failures; someone still needs to understand and own the system.
A working demo is not the same as a reliable prod app.
Code quality isn't just formatting. It's being able to understand behavior, isolate failures, reason about scaling and assess security.
Those requirements don't disappear because an agent wrote the implementation.
#CodeQuality#AIAgents
“Delegate it to an AI agent” is not a substitute for prioritization.
Agree on scope and acceptance criteria, then make time to validate the result.
Companies investing in agents should also invest in the engineers expected to supervise them.
#AIAgents#EngineeringLeadership
By introducing stricter rules, they can create an artificial delay and attribute slower progress to regulation. We already see signs of this. New models such as GPT-6 Astra appear to rely on increasingly sophisticated harnesses to achieve AGI-level benchmark results.
I believe AI leaders are pushing for an industry-wide agreement on stricter regulations for AI agents because upcoming models such as GPT-7+ may not represent the major leap people expect over today’s systems.
https://t.co/BrqJBlysg8
#AI#AIRegulation
us: we are struggling to figure out the best way to use coding agents, we don't have clarity yet
everyone else: our team is moving at speeds unheard of, all our PRs are ai generated, we've cleared 6 years of backlog
man we must really suck huh