AI providers will ban you for:
- using too many tokens
- reverse engineering apps
- trash talking their LLM
- asking for security tricks
- trying to compete with some companies
- having opposite political opinions
- directing art deemed "offensive"
Picking winners & losers
How will AI providers drag so much electricity that it can power all white-collar jobs in the world? They'll could:
- reserve service to US only
- heavily price outside US
- open data centers worldwide
- ... (what else?)
People reinventing the wheel with AI everyday now. I wonder what will happen when they discover Github... 🤯
Remember that AI providers already decided you are not allowed to work on security, they put guardrails in place so you can't ask for security breaches. You think they will not decide what other tasks you are allowed to perform with THEIR AI? Think twice.
Software is not dead. People will always look for industry standards before rolling their own version with all the risks it entails (precision, hallucinations, traceability, completeness, education, ...). They pay for trust, not code. Low-risk softs are already free with some ads
Unpopular opinion: AI makes Software Certifications more valuable. Just like you don't want random people build cars from parts and risk others’ lives, you cannot trust an AI to replace a SaaS because "it looks good" after pushing 2 buttons.
In 2006, 1 SMS was worth $0.15.
At 160 characters (~40 tokens) it means SMS were priced at $3,750/M Tokens.
Today SMS is so commoditized nobody even bothers checking its price, it is even shipped for FREE on most plans.
Probably safe to think AI will follow the same path.
Here's the narrative I keep seeing every day: "AI can code so coding is a solved problem, so devs are cooked".
But AI can read too. Does it mean we should stop learning how to read then? 😂 will books be obsolete in 2 years because Claude will read them instead of you?
Hilarious
Here's why SWE are still relevant in the Claude era: when non-tech people found out about AI, we were already prompting it. When they found out about prompting, we were already context-engineering. When they found that out, we built agents. Then, we built harnesses, etc
Claude now empowers PMs to code their own dreams. That's sweet.
Now, imagine what an actual experienced SWE can do with Claude.
Do you think we ask to build an app and call it a day? lol
In French we say: "On apprend pas aux vieux singes à faire la grimace."
Pydantic & TypeScript folks: "we have added stricter definitions, documentation, linting & pre-runtime checks to high-level programming languages, for better software quality and team collaboration"
Vibe Coders: "ok create the app"
LLMs don't translate Language to App: they translate Language to Code, which forms an App. This is an incredibly important nuance: LLMs are just the highest possible programming language, higher than Python.
LLMs abstract away cognition into probabilistic distributions (in the form of text for now). If you need deterministic outputs in your workflow, LLMs are probably not the right tool for that job. In my experience, AI should only replace ~20% of your workflows. Not 100% !!!
E.g Payment Processing
Don't give credit card details to Claude to have it process a payment end-to-end. Could it do it? Sure! But how many steps could it fail though? Instead, use the LLM on the 1 or 2 steps where "not sure"/"maybe" is a feature, like fraud detection.