HN user
Flash versions were often ultra competitive and their best in the range along with openai mini models. Always been gemini most useable and best model with nano b. Frontier is much more competitive. Anthropic haiku is like 2025 flash…
everything in LLM world is that right now.
maybe he meant about server-side features? otherwise its dumb. OAI and Anthropic are using azure/aws/gcp...
it is obvious at this stage that most of the gains are in distillation post training and having good RL simulations. the moat of the private labs are just their capability of stealing open research and data and locking the few bit they innovate on the top of it.
It will work until they IPO.
First three paragraphs and I can tell its opus 4.8
They all do that. Ask leading models about israel palestine conflict for instance
russia was indeed the ones who pushed germany to be reliant on their energy instead of going for nuclear. they funded green party to do that. and germany on their side then voted laws and pushed eu to remove nuclear from green energies list so that france wouldnt have an edge against germany shooting themselves in the foot
the eu has always been an instrument of american imperialism. leader like ursula was casted away of german politics for corruption and most of the other big names had ties to american companies like goldman sachs or’other financial institution. the eu is a prison for all of us. for a moment germany thought they could use it as an instrument to win and crush its biggest competitors (france and uk) but now they dont have an energy sector (lost thanks to their dear american friend bombing nordstream and foreign countries financing an anti nuclear narrative) and as such they now also lost the heart of their economy : their industry. the final nail in the coffin is spain opening the gates to millions of mens from less developed countries while major european economies have record youth unemployment.
its a crime against what was not so long ago some of the greatest nations on earth. now were as citizen are living under a distopia of urss with the worst of capitalism combined with the worst of communism. mass surveillance, removal of all personal freedom (freedom of speech, right to own property and cars, right to inherit, right to have a nation for our people, harshnpunishment for any contestation’up to jail timz for memes while at the same time very lenient justice toward murderers, rapists and other criminals.)
we gave away our right to exist and be nations and we did that without even a fight
that will be my last gale console
I guess the symptoms might be super hard to identify. llike a cognitive decline on some functions.
if the worms are in the brain the man is likely not fine and won't ever be like before
why use agentcore runtime then
they even published papers where they distill gpt models…
https://alignment.anthropic.com/2025/subliminal-learning/
anthropic is really adversarial.
no its not. its about mass surveillance and mass population control. regulating, watching, punishing and jailing the taxpayers while letting millions of people enter illegally and destroy the society from both sides.
europe is huxley nightmarish utopia worst parts without any of the bright ones.
well they did it against russia RTnews and even today russian players are treated like trash in Europe. just look at tennis for instance.
do they mention hwo they meddled? there is a scenario where such firm would expect mamdani elected as a favorable outcome.
Why would you need a pm or a biz analyst
4k screen is minimum for dev work and text clarity should be priority above any other feature. this will help a lot over the years
I guess he meant nova models? Which are definitely not even on the same table than thsoe models so its irrelevant
markdown became the language I hate the most thank to LLMs and specs-driven approach. everything feels so dumb right now in agentic coding. looping blindlessly and aimlessly until it compiles then until the playwright server or whatever devtools shows that it somehow works. push the code, have a llm autoreview/autofix,push to prod, run a mythos (perfect name) to identify the bug that opus 4.7 create. loops on loops on loops of some kind of zombie processes running to a "goal" that everyone seems to mystify in talks to just hide the fact that we do nothing anymore. the bottleneck never was code. it was the gate that was keeping away the Elizabeth Holmes and SBF from software engineering and it just opened.
my smell never recovered fully from first waves of covid. I am at 70%. ill try this.
nothing is really fun in trying to hack around a model to make him do things you want knowing that its barely patching it until the next disease shows up. its painful.also it is probably a temporary step before everything close up behind closed software and closed API (like anthropic tries to) and the harness wont be yours it will be on a remote cloud server. then you will just spend the time review llm generated code. also the scale and ambition of a LLM generated piece of code can be daunting. the most basic pattern since LLM can code well is overdoing. docs are longer, blogs are longer, ppt are longer, PR are longer while quality naturally drop because more people can produce this kind of content now
To experienced devs. Why are you enjoying vibecoding? It is soulless and the job is really becoming unbearable having to discuss with an agent that alternates between “good at” and totally consensual and low iq.it feels miserable to have a few turns then seeing the tool Became so bad and so low iq that you stuck in doing things garder than if you did it by yourself
I get the conversational aspect and value of it I just dont get people saying “I dont code anymore I manage agents” - besides obviously people selling AI
Taxation, admin hell, government involvement led by highly uncompetent unelected people that got to make decision on the sovereignty and future of Companies. The only chance for mistral is to escape the grasp of france and its low iq visionless political “elites” or else theyll endup like dailymotion.
Building a company in france and europe is hell even mistral ceo said this a few days ago in front of french officials
Its not ridiculous but accurate. You right the scientist generation is huge from europe but many leave europe… Brain drain in france is huge nowadays because of what op states.
And if you look at mistral biggest customers it would be lying not to say they are done through political ties. No shame in saying that. US gov and agencies created FAANG through those same mecanism
its not that loose of a version. its the reality and as probably is surely a focus of a dedicated post training RL-ing these kind of githubs. of course you would train specifically on the task. you would mix this eval data with others in thousands of githubs repos.
that is for sure what everyone does. also they train on evals with the datasets that they would be bench against.
there is no digital sovereignty in european cloud. first they would bow down to any bigger instance of power that ask for their data and being smaller companies than azure aws gcp they wouldnt have the firepower to fight back against governments. Like this one : https://www.theregister.com/off-prem/2025/11/27/canadian-dat...
second, europe has the most digitally agressive roadmap in the democratic world right now. they plan to ban vpn, enforce agressive data laws that give full power to authorities and gov to extract legally your data from your "sovereign cloud", remove anonymity from the web, enforce a cashless distopia where they can track everything and block you from using your own cash, punish you with laws against hate speech where governements decide what they define as hate speech depending on who is in power.
finally for his choices. Mistral while riding the european sovereignity wave is in fact an american owned company with european founders and the french gov trying to kill anything that they dont like touching Mistral.
OVH while a good company is definitly not providing US cloud-level data resiliency and recent events are pretty worrisome from data loss fire and hacks on customer data
Proton, maybe the only company that ever looked for sovereignity is thinking about leaving switzerland due to these opressive laws.https://www.techradar.com/vpn/vpn-privacy-security/we-would-...
also he kept the only company that is vibecoding in prod (cloudflare) and proud of it while laying off people based on the ai-religion.
It is like he made all the wrong choice if his goal was like he says to own his data and know "where is the data"
That is a summarize my wiki. Nice search feature.