Dan Walmsley

@frommyvantage7 @iruletheworldmo Do you condone violence or not? Personally I think there’s enough FUD in our media ecosystem that anyone resorting to violence based on what they read about someone online is the real problem.

@iruletheworldmo Didn’t he most probably kill the Indian kid. Lie about making it open source. How can you trust a liar. I’m not condoning violence. I’m just pointing out that we have a liar in charge of a very dangerous machine. 🤷‍♂️what’s your thoughts not Elon

please don’t hurt people under the guise of saving people. it doesn’t make you a martyr, just a gullible and cruel fool. sam is a great man, i wish him and his family and friends well. there’s enough going on and perhaps we can pull together under our shared values of non violence, democracy and human flourishing. i call for unity, not division.

@XFreeze This seems like hype designed to dethrone the company with the best coding model. I don’t think you know what you’re talking about.

I am telling you, Anthropic is not building an AI that is here to protect humanity, but rather an AI that will turn against it They're deep into the woke rabbit hole and they do not care about morality Their AI safety team is a joke. Their moral guide for AI is literally a leftist lunatic with a twisted understanding of reality, who doesn't really care if the end goal causes humanity's end They have proved this at every opportunity they've had, and it's crystal clear

@killerstorm @karpathy The slower your project moves, the more likely it is to be compromised. Evolve or die.

@karpathy Yeah, it's rather sad devs now believe that libs require weekly updates now. Nobody is even trying to write software in a way that it can be "done" - i.e. code is final, it does what it says it does. Then dependency won't be a problem

Software horror: litellm PyPI supply chain attack. Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords. LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm. Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks. Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages. Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

@Liamjsm Yes but existing (closed source) software can’t be mutated at will to do anything you want. It’s an almost infinite difference in capability when looked at that way. If

Lotta people using Claude to build software that already exists

@JadeCole2112 There are benefits to doing less precise work but at massive scale. It’s not because computers are smarter (but they will be). It’s that you can make them in factories and share one massive computer between billions of people simultaneously and it works endlessly and repeatably.

Is it just me, or are smart people spending an inordinate amount of time trying to get this stuff to do what they themselves have been doing for years?

@haha_girrrl The main problem is that you can’t verify the credential unless it is also a communication channel. if you do any kind of 2fa it does more harm than good, it’s just busywork for the user. This is why everyone moved their existing username field to “screen name”

Hot take: The username field is pointless. Change my mind.

@DanielW_Kiwi @vaxryy I’ve met some believers who, when pressed, will say “yeah most of the Bible is probably made up from folk tales” and the occasional atheist who, when pressed, will say “well… I believe there’s something mystical about why the universe exists” Shades of grey…

@vaxryy What about belief in a god? CHECKMATE ATHEISTS

everything is allowed if you are atheist

@atmoio Just because something is easy or obvious doesn’t mean it’s not also the best way to do things. He’s not an idiot - he knows it’s just prompts. What he’s doing is saying “now anyone can be a CEO”, not “I’m a super genius for thinking of this”.

AI is making CEOs delusional

Stop building agents. Start building agent coordinator agents. An agent that spawns other actions on a heartbeat, running at the core of your process or company. Trust me. And it can just be a cron job or Claude /loop!

@iamyesyouareno Ok cool. We all have to move to, say, Austria or be persecuted out of existence. We end up crammed in there like Palestinians or Jews in the Middle East, even though there’s far more of us than can reasonably fit and very few of us speak german. Does that sound like fun?

Tucker: "Every other group in the world has a right to its own homeland except White people? What? Explain how that makes sense.” White people deserve a homeland.

@MissyIsMaga @RealJamesWoods The intellectual bar for creating charts is continuously lowered until even this opinion can be represented visually.

@RealJamesWoods Well… if I had to pick one, I would definitely pick the one on the left

I am done with the Republican Party. Between this and Thune’s refusal to pass the SAVE Act, I’m done with these uniparty traitors. I’m changing my party affiliation to Independent. No wonder President Trump is fighting an uphill battle every day.

@2chancewilkins @DaveShapi Uhhhh… Dario seems pretty legit. Not sure where you’re getting this “evil” claim from, nor how it’s helpful in a policy debate. Policy assumes no intentions on the part of the actors. He’s just a guy having some serious communications issues with the US dept of war.

Let it all come out. No holding back. Let's go.

@mwilcox I vibe coded a 10 million line healthcare app and for some reason I’m still finding bugs. I blame Claude.

@StevenCCrowder @amitisinvesting @sama To be fair, Elon said the same kind of things about Dario.

@amitisinvesting @sama Altman is not a man to trusted at all just like Elon called out.

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.

@cybrdean1311 @gdb Wow it really has become a kind of movement with a life of its own now.

Even if it is, there’s almost no reason to use it over alternative options. The one thing ChatGPT had over other models was understanding conversation, and their model’s personability. 4o is still the best model that ever did that, and they canned it for “safety”. More and more people are leaving. OpenAI is a sinking ship. Switch while you can. #keep4o

it’s a good model

@elonmusk Prosecutions require discovery…

And another one down. But still no arrests …

Performing stand-up this Friday the 13th at Nevada Theater - last day for cheap tickets!

@GalatanAlin @phokarlsson Not always the best strategy. Usually early on the best approach is the one from which you learn the most.

@phokarlsson As a mathematician and later as a software engineer, I always had the philosophy of “always attack the hardest part first”. I think it is equivalent.

Christopher Alexander has an observation about problem solving that I like: you should always be focusing on solving the part that has the fewest degrees of freedom. When figuring out how to design a kitchen, for instance, there are a bunch of subproblems to solve: where to put the stove and the windows and the kitchen table. And which of these have the fewest degrees of freedom? The windows. If you want good light, there is going to be only one wall where you can place the windows, and at best two spots on that wall where the window looks natural. So you put the window there. And now what? The kitchen table, because you want to have that where the good light falls. The stove can wait because that can sit nearly anywhere. If you start by placing the stove, there is a big risk that you block the only good position for one of the other subproblems that have fewer degrees of freedom, and so the whole design will suffer.

@NCorvus58947 @KylePlantEmoji @dalepartridge @grok please confirm the words and claims of Minnesota police chiefs regarding racial profiling of their off duty officers by ICE and DHS, and whether any evidence exists that all the targeted officers were “illegal” non citizens.

@danwalmsley @KylePlantEmoji @dalepartridge False, the officer was an illegal if it's the same one I'm thinking of.

I have a friend who lives in Minneapolis. He told me the vast majority of protestors are white liberals. Not Muslims. Not Hindus. Not Hispanics. Not blacks. White liberals. They are fighting for foreign illegals who hate them. It truly is suicidal empathy.

@JackPosobiec Cool but how about just sharing it with your fellow Americans, then?

We are not surrendering the United States of America to you Bolsheviks