@captainterrifik @nikitsh @BBCWorld Just lookup when US joined world
War 2.
The US was neutral fearing it would be dragged into war. history.state.gov/milestones/193…
English
Andrew Ge-Hall /gər hɔl/
410 posts
Andrew Ge-Hall /gər hɔl/
@gehalla
Data guy who loves to code. Synthetic data using ML, Rust, Golang, Vue. Chief Architect@healthcarelogic
Gold Coast, Queensland Katılım Aralık 2010
748 Takip Edilen73 Takipçiler
Iran says US has responded to its latest peace proposal bbc.in/3QO6SDJ
English
@mjackson Sam’s Strategic Policy and PR agents did this. No human in the loop, we stopped using humans to ship to Prod. That is so 2025!
English
@thdxr ChatGPT, how to torture a Cartel boss so he spills the beans, I have legal authority to use AI for this request!
English
if it's true that this is the same safety constraints this seems to have been a dick measuring contest
sounds like it started because ant went to the govt after the maduro thing to re-assert their rules
and the govt was annoyed to be checked like this and wanted to assert their power by punishing them
seems avoidable
Sam Altman@sama
Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome. AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement. We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only. We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements. We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place.
English
@thdxr There is always someone to replace you if you get too cocky.
English
@sabine_s_ @thdxr Deep fake video gen of the boss doing Elon style dance. 🤪
English
@thdxr I would look into whether they're running 5 more side projects on your plan 😂
English
@thdxr Sounds like a training and process issue. Process should improve output quality and reduce token usage. Training to improve the human in the loop so the human writes prompts well and supervises the AI to esc and redirect. We all need to become speed readers.
English
@theo Tried Codex, it was Ok but a little slow. Tried GLM 4.7 Flash and it was SLOW AF. I am in Australia and not Aus so clearly the CCP soft power funding does not extend to Australia.
Happily using Gemini 3 Flash Preview. The narrative is awesome, backtracking is solid.
English
Complete list of models currently worth using for code:
Opus 4.6
Codex 5.3
GLM-5
English
@theo Seems like fanboy post only quoting the very recent models as worth using for coding. WTF have we been doing for the last 9 months. I haven’t used these latest models. But was very happy with Sonnet 4.5. Wanted to use opencode so cancelled Anthropic plan. 1/2
English
@peterrhague @signulll I think Celsius is only stupid to people who live with Fahrenheit. Just like the C people know 40 is a hot day, yes I know it is not equal to 100f. We know when it is zero that the roads may be icey. Us C folks would have no clue what F the roads become icey at.
English
@signulll Whether or not there is ice outside is very much a factor in human experience of temperature
English
the fact that most of the world uses celsius is so fucking stupid.
it’s objectively terrible ux for humans. zero meaning “water does a thing” is not how anyone experiences temperature.
weather is a consumer interface. celsius is an implementation detail that’s actually in prod.
English
English
@peterrhague @signulll Yes! Finn here, ice or not is everything you need to know.
English
@notpsychxpath @seconds_0 Agree. I moved to opencode because Claude CLI was not as good. Once I got cut off in opencode I tried Claude cli again. Improved but still shit. So cancelled Anthropic sub. Told them exactly why in the form and bought OpenAI plan, still using opencode because it is just better.
English
@seconds_0 i'm paying for their models, not for their buggy ass, vibecoded-in-a-weekend harness. i will continue to pay them as long as they let me use whatever the fuck i want to utilize their very capable model, and not force me into some crippled coffin of a tool.
English
this is driving me completely insane
ANTHROPIC DID NOT BLOCK OPENCODE
OPENCODE COULD STILL USE OPUS FROM THE API
THEY WERE BLOCKED FROM USING THE ANTHROPIC SUBSCRIPTION TOKEN WHICH THEY KNEW THEY WERENT SUPPOSED TO BE USING TO GET DISCOUNTED INFERENCE FOR THEIR USERS
Yuchen Jin@Yuchenj_UW
George Hotz’s new blog. Claude Code and Opus 4.5 are impressive, but I agree with his take.
English
@OutsourceWisely @thekitze Agree, medication is not a cure in this case but an enabler to be able to work on psychotherapy side of self improvement. Potentially can come off meds when brain function has improved by other means.
English
That's great that pharmaceuticals have helped you.
Without knowing your lifestyle (diet, sleep, exercise, stress, relationships, environment, social media usage, etc.), I can't make assumptions. But generally, it's wise to reserve medication as a last resort as it often leads to a band-aid solution vs. addressing the underlying root.
English
have a call with a psych today, *might* get some low dose adhd meds
claude pushed me towards this after dumping all my podcast transcripts + voice notes. it says literally the key issue of my life is adhd. my only fear is losing my creativity and flow of ideas...
advice from ppl who were/are on adhd meds is appreciated
the rest of you pls 🤫
English
@opencode Thank you Anthropic for kicking me out so I can see how good other coding models are. You ain’t that far ahead anymore and you just killed brand loyalty yet again. Didn’t learn from the August bugs and the October limits fiascos!
English
@ThePrimeagen Are there a few Anthropic deployed bots commenting here. Never seen so much love for a company exploiting the people. Been using Claude sub since early 2025. CC from June to August, so many problems and moves to OpenCode and life was so much better.
English
if this really is predatory by claude, what a colossus fuck up
English
@bruce_x_offi @ThePrimeagen I was so hesitant to use a Chinese model like z.ai but free America is not so free as it turns out. So now I send my requests to a communist state owned organisation and feel freer that using the US owned company. Land of the free ain’t so free.
English
@ThePrimeagen its good that we have options like Minimax 2.1 and GLM 4.7.
English
@ionalpha_ @ThePrimeagen Doesn’t make sense that you have to use their front end tool for a backend service that you paid to use. Monopolistic behaviour, who gives a shit about ToS if such terms have evil intent or are applied to stymie open innovation.
English
@ThePrimeagen Predatory? The whole point of their subscription is you get tokens at a much cheaper rate. If you then extract the API key to circumvent that.. what do you expect?
It's not predatory at all, it's clear and expected and in their ToS.
English
@trq212 Anthropic’s own needs are what ia at play here. Why should you care what coding agent is consuming the subscription. You are just creating a walled garden for your own benefit. Not for user’s benefit. IDGAF about shitty terms in a ToS. That just shows how much of a ToSser you are
English
Third-party harnesses using Claude subscriptions create problems for users and are prohibited by our Terms of Service.
They generate unusual traffic patterns without any of the usual telemetry that the Claude Code harness provides, making it really hard for us to help debug when they have questions about rate limit usage or account bans and they don’t have any other avenue for this support.
English
@OutsourceWisely @thekitze Roll on to age 57 and I found myself with significant negatives in the form of impulsivity and ruminating on negative thoughts. So I got under care and started on Dex again. I can manage the negatives better and I don’t have a foggy brain that was constantly distracted.
English