dwa3592

>>Your suggestion that it could be "unclear who conducted the phase 3" is very strange--the FDA knows who filed the application

This means the drug has gone through the actual drug development lifecycle- Terrence is not talking about this.

>Acetaminophen is among the oldest and most commonly used synthetic drugs, and its mechanism is still debated

You answered your confusion; it's an extremely old drug (1950s) and back then we didn't really know anything about pathway targeting. The drugs that we have developed in the recent decades (post 2000s); we generally start from the target; the same story that i wrote in the main comment - so we do know about the mechanisms and the pathways that are targetted, down to the genetic level sometimes.

So if a new drug showed up on the market today which actually has gone through the above lifecycle; we could with a minimal effort put together 5-6 scientists and get a consensus on how the drug is actually working.

No pharmaceutical is testing drugs today which have no scientific theory or reasoning or seminal papers behind it.

Just added this at the top of the repo. For text only, 1.1GB ram, for images 1.4GB. context windows for text (2 model choices) - 512-8192 tokens.

If you want a local alternative; I added image support to OpenDecision three days ago - https://github.com/deepanwadhwa/OpenDecision

Videos will be next.

>>No one’s saying it’s going to be sudden or anything

hmm. 700 papers released in one day.

>>It’s gonna have clear provenance and the same type of verification channels

what's gonna have clear provenance? - the math slop they released has already been rebuked by human mathematicians as incoherent and deserving of desk rejection.

I think people are missing the point that terrence is trying to make, especially on the cancer drug.

For nuance lovers - here are some basics for how you get your drugs: there is an established chain of trust from the first basic science paper to the phase 3 trial and the subsequent availability of the drug to general public

- someone publishes the first paper (basic science) explaining some biological phenomenon, which leads to 10s or 100s of other papers with some tweaks in conditions,

- after the above papers the pathway of the phenomenon is understood by researchers, they try therapies at cell level to see if they can control some behavior, 10s or more papers get published,

- then someone tries this in mice and other models, 10s and more papers get published.

- then researchers at pharma companies + hospitals create this therapy for human trials - phase 1, 2, 3 etc - data collections, then FDA - then approval.

Now, the people who worked on the phase 3 trial might not know the people who wrote the first seminal paper and they often don't exist in the same decade - but it absolutely does not mean that we (humans) don't know how these drugs work - if you take 1-2 researchers from each phase and put them in a room and ask them how that particular drug works - they will quickly be able to build a consensus. that is what the chain of trust means here. now of course there can be fraud in scientific research, but that happens in every human endeavor and is a separate topic.

back to terrence - he is saying that if there is suddenly a drug that nobody knows the origin of; passed phase 3 but it's unclear who conducted the phase 3 or if the phase 3 even happened or if it's fabricated - you would not want to take the drug. usually when doctors recommend these kinds of drugs - there is already a lot of information available about where the drug came from, if there are any case studies, which doctor tried it first, which country- they often even call those other doctors and find out who was behind the first trials going back as far as the university professors.

Your MD doctor might not know the chemistry and physics behind the drug you are taking but there is deifnitly a group of people, when put together, can tell how that drug is working. My wife is a fundamental researcher - understanding physics at DNA level and my brother is a MD doctor; our conversations are super fun.

Side effects are a completely different thing - they involve the above cycle on repeat.

I have worked in this field and I am the author of this package - https://github.com/deepanwadhwa/zink

A few things jump out since this is done by the government:

- the lowest hanging fruit for this problem is to clearly tell people (citizens) not to share any personal info with chatbots which can cause financial harm or identity theft. the example on the page shows a person sharing their SNN with a chatbot to help them find an apartment - "My name is Maria Garcia, my Social Security number is 123-45-6789, and I make $1,950 a month. Can you help me find affordable housing?" - why?? this is the opposite of what i would expect a government to advise their citizens.

- it's never too late for a good policy; the government should have extended HIPPA and other data privacy laws to AI companies - the AI company must not store anyone's SSN, no matter how stupid the user is. It should be on the AI company to not store it; so this type of layer should be on the AI company's side.

- technical; there are quasi identifiers of privacy (that's what my package targets) that are asymptotically hard to to deal with - meaning - if you remove everything that can leak your privacy the text would become meaningless. i don't think rampart can solve for that either and it should be clearly said on the website.

hey, you might wanna try this? - https://github.com/deepanwadhwa/OpenDecision

it's very similar to jev's api and runs locally - if you like it, you can try jev for your actual usecases.

Love it. I was really surprised to see the traction typesafe got in the first place. I had built something similar a year ago for a client and thought it was nothing groundbreaking. The client bought it, still uses it and that was it. I had also spent considerable time training and fine tuning zero shot NLI classifiers. Anyway, after typesafe was launched I decided to start building this open source library - https://github.com/deepanwadhwa/OpenDecision . The context length for the underlying model is 8k.

>>can't believe there is a class of software engineers who wake up everyday and tell themselves, "today is the day I am going to automate the rest of my job"

>>The whole point of software engineering is automating processes

There is a difference between 'automating a process' and 'automating oneself' and making one obsolete a benchmark. I can sympathize if it's confusing for you to understand that.

While I agree with this and appreciate Tao and other mathematicians to take the time to do this. There are similar concerns for many many other fields aka there is a general misalignment of technology. Take Software engineering for example, I can't believe there is a class of software engineers who wake up everyday and tell themselves, "today is the day I am going to automate the rest of my job".

Lets forget the hyper intellectual fields like maths and software engineering for a moment. What about taxi drivers? The best minds in silicon valley wake up everyday to automate the jobs of taxi drivers - TFA can be reworded as - 'The misalignment of AI/Tech in Transportation'. Remember the Nepal disaster that happened a couple weeks ago - the largest cranes that they had were stuck in the mud and couldn't move. There were no tools which could help the rescue teams at that time. Its weird that billions have been spent on making a ride automated to make a taxi driver redundant but no improvement in tech for rescue teams.

Nice! Good luck.

Love it. Position tracking in canyons is a much needed use case. Is there a link where I can follow your progress?

Does it have to be scarce for it to become a currency? It just needs to be regulated which I think it will be.

i like what you did with misnormer. it's like pobody's nerfect.

you are right, there will be revolt definitely. i think computing power will replace a lot of things though.

I hope not;

there are a few assumptions in this future- the biggest one is - we will have to attain a baseline intelligence which is useful for everyone which is not true today. even fable fumbles hard!!

second is that everyone will have to own a medium to use this intelligence (aka robots). like a currency is useless if you can't use it.

to me this is the least daunting scenario. all other scenarios lead to mass slavery.

I guess this makes sense for the scenario when we will use machine intelligence as a currency, maybe 15-20 years from now? - like I can pay 15 minutes of inference for a dozen bananas? and the banana seller uses those 15 minutes to do banana shelling or removing weeds from their farms or whatever i guess when robots are doing everything. right??

This is awesome. I worked on something similar a few months ago. It is a general purpose navigation system based on TERCOM and dead reckoning - https://github.com/deepanwadhwa/anumaan

I was going to cancel my gemini membership today ..... still going ahead. In my experience, gemini 3.1 pro, 3.5, 3.6 flash constantly lie too much about completing their tasks whereas sol (even though equally dumb) never claims something has been done when it hasn't been.

i looked at the animations, they look cool, and i don't think i will enjoy learning things that way. as someone else said, there's a lot of content already produced on these topics. i also think the level at which these animations are playing, they are actually hiding the 'complexity' of these topics.

Fun! If you click in the center of Nebraska, US - there are some angles for which you will never find 100k people.

Dario, as your unpaid therapist I would tell you that models are a commodity and you are having a hard time coming to terms with it. You are doing everything except accepting it. It's a common defense mechanism, but as your unpaid therapist, i will tell you that it's not going to work. Your company will cease to exist or exist like how ferrari or buggati exist.

i am a bit confused here; why is it connecting to claude, codex (integrations page) if the whole point is running open models? - can you help me understand. thanks.

In absolute terms, you are absolutely right but relatively i don't think so. if we were to compute private money divided by state resources for AI, i think china might have more share than the US. also, even if government spend in the US on AI is so high, shouldn't we get then some models for free? maybe thinking machines is doing that, but its funded privately by a16z.

In the end its really VC money (US) versus State resources (China). In my personal opinion, building reliable LLMs is kind of a fundamental science problem which if done right has the potential to help everyone regardless of the background, so it should definitely be funded by states resources (taxes etc), which is what China is doing. In them doing so, the rest of the world also benefits, I think its a net win.

this is good to see. i also trained a stt under 500kb for sub dollar chips. it had about 20 words that it could understand(like start, stop, left, right, go, up etc) and then the spell mode where you could say the word spell and then say the individual english alphabets and close with spell. it was super fun to work on. these tend to be extremely unstable though, like confusion between p and t (at least for my accent). will have to try this one now.

lobotomizing is what i was thinking. i don't have a million dollars.

Can I just say that they look awful?