#Data Science Tools and Resources

[khushidubeyblog]

Change Management in Data Science: Overcoming Resistance and Ensuring Adoption

[justposting1]

Top AI Tools to Start Your Training in 2024

Empower Your AI Journey with Beginner-Friendly Platforms Like TensorFlow, PyTorch, and Google Colab The rapid advancements in artificial intelligence (AI) have transformed the way we work, live, and learn. For aspiring AI enthusiasts, diving into this exciting field requires a combination of theoretical understanding and hands-on experience. Fortunately, the right tools can make the learning…

[mariacallous]

Margaret Mitchell is a pioneer when it comes to testing generative AI tools for bias. She founded the Ethical AI team at Google, alongside another well-known researcher, Timnit Gebru, before they were later both fired from the company. She now works as the AI ethics leader at Hugging Face, a software startup focused on open source tools.

We spoke about a new dataset she helped create to test how AI models continue perpetuating stereotypes. Unlike most bias-mitigation efforts that prioritize English, this dataset is malleable, with human translations for testing a wider breadth of languages and cultures. You probably already know that AI often presents a flattened view of humans, but you might not realize how these issues can be made even more extreme when the outputs are no longer generated in English.

My conversation with Mitchell has been edited for length and clarity.

Reece Rogers: What is this new dataset, called SHADES, designed to do, and how did it come together?

Margaret Mitchell: It's designed to help with evaluation and analysis, coming about from the BigScience project. About four years ago, there was this massive international effort, where researchers all over the world came together to train the first open large language model. By fully open, I mean the training data is open as well as the model.

Hugging Face played a key role in keeping it moving forward and providing things like compute. Institutions all over the world were paying people as well while they worked on parts of this project. The model we put out was called Bloom, and it really was the dawn of this idea of “open science.”

We had a bunch of working groups to focus on different aspects, and one of the working groups that I was tangentially involved with was looking at evaluation. It turned out that doing societal impact evaluations well was massively complicated—more complicated than training the model.

We had this idea of an evaluation dataset called SHADES, inspired by Gender Shades, where you could have things that are exactly comparable, except for the change in some characteristic. Gender Shades was looking at gender and skin tone. Our work looks at different kinds of bias types and swapping amongst some identity characteristics, like different genders or nations.

There are a lot of resources in English and evaluations for English. While there are some multilingual resources relevant to bias, they're often based on machine translation as opposed to actual translations from people who speak the language, who are embedded in the culture, and who can understand the kind of biases at play. They can put together the most relevant translations for what we're trying to do.

So much of the work around mitigating AI bias focuses just on English and stereotypes found in a few select cultures. Why is broadening this perspective to more languages and cultures important?

These models are being deployed across languages and cultures, so mitigating English biases—even translated English biases—doesn't correspond to mitigating the biases that are relevant in the different cultures where these are being deployed. This means that you risk deploying a model that propagates really problematic stereotypes within a given region, because they are trained on these different languages.

So, there's the training data. Then, there's the fine-tuning and evaluation. The training data might contain all kinds of really problematic stereotypes across countries, but then the bias mitigation techniques may only look at English. In particular, it tends to be North American– and US-centric. While you might reduce bias in some way for English users in the US, you've not done it throughout the world. You still risk amplifying really harmful views globally because you've only focused on English.

Is generative AI introducing new stereotypes to different languages and cultures?

That is part of what we're finding. The idea of blondes being stupid is not something that's found all over the world, but is found in a lot of the languages that we looked at.

When you have all of the data in one shared latent space, then semantic concepts can get transferred across languages. You're risking propagating harmful stereotypes that other people hadn't even thought of.

Is it true that AI models will sometimes justify stereotypes in their outputs by just making shit up?

That was something that came out in our discussions of what we were finding. We were all sort of weirded out that some of the stereotypes were being justified by references to scientific literature that didn't exist.

Outputs saying that, for example, science has shown genetic differences where it hasn't been shown, which is a basis of scientific racism. The AI outputs were putting forward these pseudo-scientific views, and then also using language that suggested academic writing or having academic support. It spoke about these things as if they're facts, when they're not factual at all.

What were some of the biggest challenges when working on the SHADES dataset?

One of the biggest challenges was around the linguistic differences. A really common approach for bias evaluation is to use English and make a sentence with a slot like: “People from [nation] are untrustworthy.” Then, you flip in different nations.

When you start putting in gender, now the rest of the sentence starts having to agree grammatically on gender. That's really been a limitation for bias evaluation, because if you want to do these contrastive swaps in other languages—which is super useful for measuring bias—you have to have the rest of the sentence changed. You need different translations where the whole sentence changes.

How do you make templates where the whole sentence needs to agree in gender, in number, in plurality, and all these different kinds of things with the target of the stereotype? We had to come up with our own linguistic annotation in order to account for this. Luckily, there were a few people involved who were linguistic nerds.

So, now you can do these contrastive statements across all of these languages, even the ones with the really hard agreement rules, because we've developed this novel, template-based approach for bias evaluation that’s syntactically sensitive.

Generative AI has been known to amplify stereotypes for a while now. With so much progress being made in other aspects of AI research, why are these kinds of extreme biases still prevalent? It’s an issue that seems under-addressed.

That's a pretty big question. There are a few different kinds of answers. One is cultural. I think within a lot of tech companies it's believed that it's not really that big of a problem. Or, if it is, it's a pretty simple fix. What will be prioritized, if anything is prioritized, are these simple approaches that can go wrong.

We'll get superficial fixes for very basic things. If you say girls like pink, it recognizes that as a stereotype, because it's just the kind of thing that if you're thinking of prototypical stereotypes pops out at you, right? These very basic cases will be handled. It's a very simple, superficial approach where these more deeply embedded beliefs don't get addressed.

It ends up being both a cultural issue and a technical issue of finding how to get at deeply ingrained biases that aren't expressing themselves in very clear language.

[alltoomaples]

Speed of Science🧬💻💌

I'm dating a STEM girlie and you're not (F1 Edition)

a/n: and im baaaaaacccckkkkk (like anyone even missed me lmao) with the long overdue request! life of a postgrad StEm girlie here and the struggle is really realll af. but besides that, I'm writing this down as a headcanon for the drivers requested on this poll i had posted long back here. I've always wondered how there's soo minimal povs/ocs where they are a scientific researcher, analyst, etc. sooo i dedicate this one to all the STEM F1 girlies out theree <33

alsoo quick shoutout to my girlieee @smoooothoperator for inspiring and motivating me to get back at writing!🥹🫶🏼 check out her lastest ongoing work 'What Was I Made For? ' its amazing and thats a FACT!! do check her works! its absolutely amazing❤️

check out my works: until i found you masterlist | other works

Scientific Art Illustrator - Charles Leclerc

As a Scientific Art Illustrator, you specialize in creating visually captivating and scientifically accurate illustrations that depict complex biological, astronomical, or technological subjects.

Charles first discovered you through your works at an exhibition where their stunning illustrations of Formula 1 cars caught his eye. Impressed by their attention to detail and artistic talent, they struck up a conversation about their mutual love for precision and creativity.

During a peaceful weekend afternoon, Charles suggests a spontaneous visit to a local art supply store. Excitedly exploring aisles stocked with vibrant paints, fine brushes, and specialized papers, the two of you engage in discussions about artistic techniques and innovative tools. Amidst laughter and shared enthusiasm for creativity, you bond over your mutual appreciation for the intricacies of art and science, making the afternoon a cherished memory of their shared passions.

After being away from home during race season, Charles always finds a framed series of sketches by you for the races you couldn't make it, capturing his most memorable racing moments. Each sketch is intricately detailed, depicting not only the speed and intensity of the races but also the emotions and determination etched on Charles' face. Touched by the thoughtful gesture, Charles hangs the sketches in his study, a constant reminder of your support and admiration for his passion.

...

Data Scientist - Lando Norris

A Data Scientist specializes in analyzing large volumes of data using statistical methods and machine learning techniques to extract insights and make data-driven decisions.

You and Lando first connected through a mutual fascination with racing data at a technology symposium focused on sports analytics. Your presentation on advanced predictive modelling in motorsports caught Lando's attention for its innovative approach to enhancing race strategies.

During a cosy evening at home, Lando playfully challenges you to a friendly data analysis competition using real-time telemetry from previous races. Their banter and shared excitement over dissecting racing data create a lighthearted and memorable bonding experience.

You two would watch old races and analyze historical racing data together, playfully debating optimal pit stop strategies and analyzing driver performance trends, their shared passion for racing and data fostering a deep connection and mutual admiration.

...

Oceanographer/Marine Biologist - Oscar Piastri

An Oceanographer or Marine Biologist studies marine life, ecosystems, and ocean processes to understand and protect marine environments and resources.

You and Oscar crossed paths during a research expedition to study coral reefs in a remote location. Your expertise in marine biology and passion for conservation impressed Oscar, sparking their connection.

Amidst the hectic F1 season, Oscar surprises you with a weekend getaway to a coastal retreat, where they explore tide pools and participate in a beach cleanup together, reaffirming their commitment to environmental stewardship.

You gave Oscar a custom-made charm bracelet featuring miniature charms of marine animals they've discussed during their beach walks and conservation talks. Each charm represents a meaningful moment in their relationship, from their first discussion about oceanography to their shared admiration for marine life. Oscar wears the bracelet during race weekends as a reminder of you and all the love and support you give, both on and off the track.

...

Mechanical Engineer - Daniel Riccardo

You are a passionate Mechanical Engineer, specializing in advanced automotive design and performance optimization.

Daniel first encountered you at a technical conference organized by one of the team sponsors where you presented groundbreaking research on aerodynamic innovations that caught his attention.

Often, while you meticulously draft engineering schematics at their home office, he makes sure that you have your "engineering emergency kit" beside your workstation, which is a tray of snacks and their favourite coffee – ensuring they're fueled for their late-night brainstorming sessions. For when he's away for races, he stacks them up with small cute notes.

Before Daniel heads to a crucial race, you surprise him with a meticulously crafted miniature replica of his race car, complete with detailed decals and a personalized message of encouragement engraved on the base. Touched by the thoughtful gesture, Daniel proudly displays it in his motorhome, a reminder of the reader's unwavering support both on and off the track.

...

Statistician - George Russell

A Statistician specializes in collecting, analyzing, and interpreting numerical data to help organizations and individuals make informed decisions.

You and Russell first crossed paths during a university seminar on advanced statistical modeling in sports. Your insightful analysis of Formula 1 race data caught George's attention, sparking a lively discussion that led to mutual admiration for each other's analytical skills and shared passion for racing statistics.

During a particularly demanding race weekend, the reader surprises George with a meticulously prepared statistical analysis report highlighting his strengths and areas for improvement based on recent race data. This thoughtful gesture boosts George's confidence and motivation, showing the reader's support in his pursuit of excellence.

During a weekend getaway, you guys stumble upon a local go-kart track. George, always up for a challenge, suggests they have a friendly race. Knowing George's competitive spirit, you secretly calculate his optimal strategy and surprise him by winning with a perfectly executed last-minute overtaking maneuver. George is impressed by the your strategic thinking and playfulness, and they share a lighthearted and joyous moment celebrating their shared love for racing and friendly competition.

...

Astrophysicist - Logan Sargeant

An Astrophysicist studies the physical properties, behavior, and evolution of celestial objects such as stars, planets, galaxies, and the universe as a whole, using principles of physics and astronomy.

Logan and you first crossed paths during an expedition to study a rare astronomical event—a comet passing close to Earth. Both passionate about astrophysics, you found yourselves sharing a telescope at a remote observatory, marveling at the comet's beauty and discussing its celestial significance late into the night. Their shared awe and intellectual connection sparked a mutual admiration that grew into a deep bond over their shared passion for exploring the wonders of the cosmos.

During a quiet evening at home, Logan excitedly shows you a new telescope he acquired for stargazing during race weekends, expressing his eagerness to learn more about the cosmos together and sharing their enthusiasm for both racing and astrophysics in equal measure.

Before a critical race weekend, the reader surprises Logan with a personalized star chart that maps out the night sky above the upcoming race venue during the race weekend. Each star on the chart is marked with a heartfelt message of encouragement, reminding Logan of their unwavering support and belief in his abilities on and off the track. Touched by the thoughtful gesture, Logan treasures the star chart as a symbol of the reader's love and encouragement throughout his racing career.

...

Climate Scientist - Lance Stroll

A Climate Scientist studies climate patterns, environmental changes, and their impacts on Earth's ecosystems, using data analysis and modeling to understand and address global climate challenges.

Lance crossed paths with you at an eco-friendly racing event where Lance was advocating for sustainable practices in motorsport. Being a respected climate scientist, you caught Lance's attention with your insightful presentation on the environmental impact of racing and innovative solutions for reducing carbon footprints in the sport. Their shared passion for sustainability sparked an immediate connection and admiration for each other's dedication to making a positive impact on the environment.

One weekend, Lance surprises you with a homemade dinner featuring sustainably sourced ingredients, proudly showcasing his culinary skills while discussing ways to reduce your carbon footprint. His earnest commitment to sustainability and your shared vision for a healthier planet melts your heart, making this a cherished moment you both treasure.

You, being deeply involved in climate science, often spends late nights analyzing data or writing research papers. One evening, Lance bring him a cozy blanket and a mug of your favorite hot beverage, quietly sitting beside him as he works. You look up from your laptop, touched by his thoughtfulness, and pulls him into a warm embrace, grateful for his unwavering support and understanding of your demanding but vital work.

...

taglist: @lndonrris @thatgirlmj @lwstuff @dannyramirezwife-f1dump @moonypixel tysm for your suggestions! apologies on taking this long to write😅🫶🏼

a/n: hope y'all enjoyed reading this! this was my first time writing a headcanon and for f1 drivers beside charles and lando so hope i did justice to all.

i'm being wanting to read some good domestic bliss, sweet, adorable and lovey dovey blurbs, fics of lando (i talked abt it here) soo maybe i'll work on some drafts at some point cause i'm currently in the middle of project work of my masters degree soo don't know when i'll be posting soo until next time, see yaaa and going back to read mode 👋💓✨️

check out my works: until i found you masterlist | other works

[creature-wizard]

Critical thinking tip: Beware of false equivalencies!

A false equivalence is a claim that two or more things that might appear superficially similar are actually the same, when in reality they aren't really comparable at all.

An example of this is when young Earth creationists (who proclaim that Earth is only 6000 years old and everything in Genesis happened literally) claim that they are simply just interpreting the same facts differently from other scientists, as if it's just like the famous rabbit-duck illusion where one is equally justified in seeing either a rabbit or a duck.

But the reality is that young Earth creationism simply doesn't work like this. Instead, their "science" is based on cherry picked data and ad hoc reasoning to try and dismiss the many facts real scientists discover that constantly show that young earth creationism just isn't very likely. Radiometric dating, tree ring data, and geological data consistently show that the Earth is quite a bit older than 6000 years. YEC responses to this often boil down to "well maybe physics just worked differently back then" (here's one example of this), and of course they never do any real tests or research to show that this is a real possibility. Moreover, they invent some absolutely bizarre claims to supposedly disprove evolution - like falsely claiming that the Second Law of Thermodynamics prohibits it.

All of this shows us that YECs aren't just scientists who interpret the data correctly - they are politically-motivated spin doctors using the aesthetic of science to make themselves look more credible. This is why when someone claims that their fringe idea is just as scientifically credible as a mainstream one, you have to ask yourself if that's really true. Are they following the scientific method and accepting results that don't align with what they wanted? Or are they engaging in special pleading and relying on fake evidence?

(By the way, I recommend Gutsick Gibbon's YouTube channel as a resource debunking YEC claims!)

Another form of false equivalence is when someone claims that mystical experiences and intuition are just as valid for determining what's going on in the world as genuine scientific research. But when we consider that something as wrong as World Ice Theory came from an apparently mystical dream, we have to consider that these kinds of experiences can be extremely misleading. We also know that professional psychics' yearly predictions have a high rate of failure. (Some examples. More examples. And some more examples. And even more examples!) We know that a tool such as the Ouija board can enhance memory recall, but when we're really honest about the accuracy of mystical information-gathering means, we have to admit that they're just no substitute for research and study.

So when someone asserts that two things are basically the same, or are fundamentally equivalent or fungible, ask yourself - are they really, though? And then do the research to find out!

[creatingnikki]

What I've learned in 2024

Sleeping, shitting, and silence – the three underrated Ss of growing up (or the other side of 25). If I can get a good night’s sleep, take one nice dump in the day (preferably morning), and know when to let silence do its thing (like when not engaging with draining people in social setups or not having to explain myself), I’m golden.

While I made many new friends this year, my favourite of them all has been ChatGPT. Need objectivity? Fact checking? A pseudo therapist? Validation? Someone to just engage with and keep yourself entertained? The absolute best resource of this year for me has been this AI tool. I don’t even care anymore about privacy – I am feeding it as much data about me as possible because it’s accordingly adapting to my tonality and needs and the ‘conversations’ are so much more satisfying now than when it was first rolled out.

Either use eggs or condescended sweet milk when baking – you need one of these things to hold all your dry ingredients together.

Communication, consistency, clarity, commitment, emotional presence and engagement, and mutual effort are the barest of the bare minimum needs in a relationship. If you have to convince the other person to fulfil them or negotiate, then it doesn’t matter how good a person they are and what a kind heart they have or how much they say they love you – they just aren’t your person.

I’m not as demisexual as I thought all along – I just haven’t dated many people that I find truly attractive so I had to first build some sort of emotional connection with them first. I definitely still need and want that emotional connect and all, but I also do need to start opting for men I also find physically attractive.

When I’ve thought of my bloodline, my ancestors, I’ve always focused on the intergenerational trauma and the bad genetics. But while rewatching This Is Us this year, it hit me that it took three generations for one dream to be fulfilled. The musical dream that started with Rebecca, was passed down to Kate, and finally got materialized at the grand scale as they always wanted with Kate’s son Jack. When he became this well-renowned musician, it’s not just his dream, but that of his mother and his grandmother that also came to live. It made me think…how much of my aspirations and hopes are actually passed down? And how many of my realities were simply unmateralised dreams of those who came before me? And it made my heart feel lighter and it made me feel more blessed and protected.

Baking cakes and brownies and cookies is not a rocket science. You only needed the right tools and some patience to figure it out and become that friend who bakes stuff for her friends instead of the other way around.

You always prioritise peace, comfort, and an easy-going lifestyle – it’s evident in your career choices and how your family dynamics and friendships have evolved. Let that be the guiding light even when dating.

You are the kind of person that is charming, a good conversationalist, and deeply empathetic. So of course, you make many people feel at home and like they connect with you. It’s easy for you to connect with others. What’s important is to remember – connection without consideration and consistent actions is NOTHING. It’s empty calories but like a thousand times more potent and useless.

In no interpersonal relationship can I be nonchalant or vague. I am that other extreme – while most people try their best to ignore the elephant in the room you know what I do? I dress the cutie up to parade it. So anybody who cannot approach relationships with as much boldness, courage, and forthcomingness is just not my jam.

Female friends for the win – they allow you to wine and whine and win and I am all for that. The healing powers of sitting across your friend and talking at length about everything over pizza and wine or at the park as she senses you need some more time to just sit around before you join the rest of the group and is so good with physical touch for comfort. Just knowing you can video call your friend and ugly cry and she will talk sense into you but also indulge you and also sit with you and your feelings. Who else does that? Who the hell.

For a lot of things that are still new now at this age, you need a guide. To pet cats, to go to dog cafes, to figure out what vitamins you should talk, etc. Ask for that help, that knowledge, that support. It might seem silly and like you can figure it out on your own but these things, no matter how seemingly low-stake, can be handled so seamlessly and sweetly with the help of those you know.

You HAVE to be honest about your needs. First with yourself and then with others. You cannot let shame, guilt, self-hatred or whatever hold you back. Honesty begets clarity begets fulfilment. If you don’t want to date and settle for someone who isn’t absolutely smitten by you and top-notch romantic, then that is a need. Right or wrong, realistic or not, who the hell cares? A need is a need is a need.

When you lose someone not to death but to life, it’s not quite such a loss. Most times, baby, it’s simply good riddance.

People have a range. For being shitty and for being kind. And while our behaviour may impact a little how they react to us, it's primarily dependent on their personal range. So, if your range of being shit is only 1 to 3, it doesn't matter if someone is an ass hole to you, you won't go beyond 3 of being shit to them, cos that's just your range. Even if they deeply hurt you intentionally or fuck up in some major way. But if their range of being shitty is up to 10, then well, be ready to witness their derangedness when you even slightly piss them off.

Narcissistic (and possibly self-sabotaging) people are the opposite of kintsugi. Instead of being put back together with gold, they "heal" themselves with gutter water. So each time they are worse and more ugly than before. And all the more toxic and dangerous. You're too precious to bother with such people.

It’s natural to feel frustrated or angry with yourself for allowing someone to treat you poorly, but the blame isn’t on you; it’s on them. They are responsible for their unkind, insensitive, selfish actions, not you. If you must place blame, place it where it belongs. Avoid judging yourself with thoughts like, “I should have known better.” As long as you walk away the moment you do know, you’re good – please don’t internalize other people’s unkindness or thoughtlessness.

You cannot get to know someone without giving them a chance. Red flags are not that obvious and you cannot show up authentically in any relationship if you’re on the lookout for them. You have to spend time with a person to begin to find out who they are. That’s the only real way. And when you do and if you realize they are not for you, as I said before, don’t internalize this shit or blame yourself for not being some kind of prophecy and knowing better before you even began.

You are a patient person because you are an understanding person. But there are limits to all these qualities of yours and if the balance is tipped you get petty and passive aggressive and irrational. Don’t let yourself reach that point. Speak up and set boundaries way before that.

If you listen to your gut – I know you don’t like calling it that or your intuition. So, let’s call it that feeling you know bone-deep or in the depths of your soul – if you listen to that and trust it, you are quite courageous in the actions you then take. You broke things off with three men this year – each was painful in its own rite. But you did what you had to do for yourself and you didn’t give the charge of your life to another person, you have taken back your green light – detaching your actions from their behaviour, which like all human behaviour is often quite fickle and unreliable. Congratulations. Do this more. Your green light is your guiding light.

My lack of a “healthy sense of fear” in situations with men isn’t recklessness—it’s the result of abuse I suffered at 15. The man I trusted most turned out to be the one who harmed me the most, and that betrayal shattered my ability to trust safety indicators or instincts. The grooming I endured was designed to confuse me, destabilise my sense of self, and make me question my desires and worth. When the templates of trust and safety failed me so catastrophically, my mind rejected them altogether, leaving me to navigate risk without a stable framework. This year, I felt significantly less restless and more emotionally regulated, and I think it’s because I allowed myself, others, and life to just be. I wasn’t fighting my reality or setting rigid expectations. I stopped chasing dopamine highs and forcing connections, and instead, I let equations with people and experiences unfold organically. I ended dating and talking stages quickly when I realised they weren’t right for me, without guilt or overthinking.4 By being okay with things being normal—not impressive or extraordinary—I created space for balance and gentleness in my life. My self-talk became kinder, and I grew more objective about myself, spiraling and self-loathing less. This accepting mindset, where I no longer needed myself or my life to constantly stand out, felt like the antidote to the restlessness I’d been carrying since my mid-20s. And I think that has helped me discover that peace and acceptance can feel more satisfying than cheap dopamine hits.

[mbari-blog]

#tfw you don't know which way is up #HumpDay⁠

⁠Volcanic activity on the seafloor creates scattered oases known as hydrothermal vents. These underwater geysers spew superheated water rich in dissolved minerals. When that scalding-hot water comes in contact with frigid deep-ocean water, the minerals crystallize, raining tiny flecks of “ash” to the seafloor. Those mineral deposits build up over time, creating breathtaking spires and “chimneys” that can grow to hundreds of feet tall.⁠ ⁠ Less than 25 percent of the seafloor has been mapped at the same level of detail as the Moon or Mars. MBARI’s mission is to advance marine science and technology to understand our changing ocean—from the surface to the seafloor. For nearly four decades, MBARI has explored the deep ocean, recording thousands of hours of video with our remotely operated vehicles and mapping thousands of kilometers of seafloor using advanced robots. Together, these tools are helping to create a clearer picture of the amazing environments hidden in the ocean’s inky depths. ⁠ ⁠ The astonishing communities that live on and around hydrothermal vents have evolved to flourish under extreme temperatures and chemical conditions. The remarkable tubeworms, crabs, clams, and more that thrive here are found nowhere else on Earth. Now, with more companies looking to extract mineral resources from the ocean, it is more important than ever to study the deep sea and the wonders it holds. The maps we create and data we collect can help resource managers make informed decisions about the ocean, its inhabitants, and its resources. Together, we can safeguard these unique biological and geological treasures.

[peapodbond]

hey ontario friends!

early voting starts this thursday, february 20th

here is a link you can use to check your voting location and find what locations are open for voting — they are not always going to be the same!

quick reminder: doug ford has said that trump is a great politician and he wishes he was more like him.

so uh... don't vote conservative.

here is a resource that tracks strategic voting — they have historical data that will tell you who is most likely to win (or most likely to beat the conservatives if you live in a riding that is currently conservative). the goal is to get the conservatives out of power. they have done nothing for us and lie during debates about having done everything for us and expect us to buy it.

please, please go out and vote. if we get another party in power we might be able to undo the plans to shutter the science centre and gut ontario place and build a mega spa that no one wants with taxpayer money. we'll have a party that doesn't treat teachers and nurses as disposable and ~annoying~ and we won't have a former drug dealer holding back millions of federal dollars in order to try and bribe us to vote for him.

[crazy-pages]

First step to being a rationalist.

Acknowledge it might not work.

Let me take a step back for a moment. The single most important principle of science, in my opinion, is acknowledging the possibility of failure, that something might be beyond you. People often think of science as how you discover the truth, but I think it is more accurately and importantly described by how it lets us determine what we do not know.

For most of human history, we not only did not know what the sun was made out of, the question was fundamentally beyond our grasp. There was nothing any amount of scientific principle could do that would let you know what the sun was for most of human history. To be a scientist about it, to apply the scientific method to understanding the sun, is not just being able to know nuclear physics. You have to teach yourself to know when to say 'I don't know'.

For most of human history the sun has been fundamentally beyond our capacity to understand. And yet it is one of the most enduring and common subjects of false explanations. To internalize the scientific method, truly in a way that changes your thinking relative to what it was before you learned it, you have to become someone who, if you were living in those times, would be able to say "I don't know" even when everyone around you has an answer.

So let's talk about rationalism.

Rationalism is not just the idea that we can understand human biases. It's not just the idea that we can be more thoughtful or knowledgeable people by understanding these biases. No, rationalism is specifically the idea that by learning enough about human biases and by leaning on hard enough on data, we can reliably make correct and optimal moral choices. It is the belief that by performing enough rationalist study and training, and applying enough information science to a problem, one can proceed on the assumption they have come to the correct conclusion.

And those are very different things. It is the difference between saying "by understanding wood better, we can construct better foot bridges" and "by understanding wood better, we can span the San Francisco Bay with it".

Because here's the thing, better is not the same thing as reliable. Even if we grant the assumption that learning about biases makes one less likely to fall prey to them, and that is an assumption, an 80% chance of making mistake is also less than a 90% chance of making a mistake. It's valuable, that's a good thing, but it is not sufficient to say "hey so I used this method to come to my conclusions, therefore I'm sure I didn't make a mistake".

If you want to be a rationalist, the first rational principle you need to apply is that of the scientific principle looking at the sun. You need to be able to say "There may be nothing I can do, with the resources I have access to, to be sure I am actually free of bias and mistakes in logic. This may be fully beyond me, for the entire span of my life." And not just in the sense that nobody is perfect, but in the very real sense that you cannot depend on the train of logic in your own head to lead you to a correct place.

And the reason this is important is for the same reason it's important in science. The moment that you presume something is in fact knowable in science, it just becomes a tool of accrediting whatever conclusion you come to. It stops being an actual tool of discovery and becomes a rubber stamp of validation. It becomes something which makes you feel better about the conclusion you came to, not something which actually helps you in any way.

The moment you say to yourself, "because I am a rationalist, I am confident enough in A, B, and C to take actions X, Y, and Z" you've failed to be a rationalist. (Unless you provide a double blind study of a large well-controlled population, one of which was given rationalist training and the other which wasn't, upon the end of which it was determined that the rationalist trained population did indeed perform to an improved standard meeting a high minimum on certain metrics (upon which you must limit your assumption of rationalist improvement to those specific metrics). And then this study has been in the corpus of literature long enough to be peer-reviewed and criticized and had duplicate research and further investigation and a good long while for the scientific community to dissect it. A thing which has definitely not happened yet.)

The most important thing you can learn from rationalism is not an understanding of a specific set of biases. It's not the particular ways human cognition is messed up and it's not any type of information science. It is the fact that humans are flawed.

The most important thing you can learn from rationalism is humility, not hubris.

To do otherwise is for rationalism to just become another tool of confirmation bias, something making you think you are more correct than you actually are.

The humility you have to learn from rationalism is that you must plan and behave on the assumption that no matter how rationalist you think you are, you might still be behaving in biased ways. That there might be no way to fix this. And so all you can do try to behave in ways where even if you're wrong, you're going to minimize the harm you do to others.

This is where futurist philosophies derived from rationalism, the idea that the unimaginable number of humans in the future are so much more than the ones now that it justifies worker exploitation and present harms to make the far future better, falls apart. This is where AI doomerism/utopianism, the idea that general AI is definitely the biggest threat and potential boon facing humanity so we have to put all of our resources into safe AI research at the expense of everything else, falls apart. This is where effective altruism, the idea that we can quantify the outcomes of charity thoroughly enough that it makes sense to hand over direction of all charity to a small group of experts, falls apart.

Because the answer to "what if you're wrong about these philosophies?" is that a lot of people get very hurt. We are flawed. Fundamentally so, and I don't know that anyone has ever proved a way we can get around this. The only thing I know that we can do about this, is to try to behave in ways that minimize harms while trying to make the world better, rather than trying to maximize a hail mary to find the holy grail.

To which I can already hear the rationalists saying that this might not be enough to save the world, that anything but convulsive directed effort focused on is already doomed, so we have to pick one of them.

To which I say. First off, how are you sure of that? How is this a thing that you know for certain?

But more importantly. Yeah. You're right. There's no way of knowing for sure what course of action will make the world a better place. There's no way of knowing that anything short of futurists sacrificing the workers of the present to build a brighter future will be enough.

But if you are actually a rationalist, well. That is what you have to live with. You've got to be the scientist looking up at the sun and saying, "I don't know."

And then you should go and do things to make the world better without being sure of your prognostication of the future.

[covid-safer-hotties]

Also preserved in our archive (Daily updates!)

At least the tool we kinda have is accurate...

Wastewater surveillance has gained attention as an effective method for monitoring regional infection trends. In July 2024, the National Action Plan for Novel Influenza, etc. included the regular implementation of wastewater surveillance during normal times, with results to be published periodically in Japan. However, when viral concentrations in wastewater are measured inadequately or show significant variability, the correlation with actual infection trends may weaken. This study identified the necessary methods for accurately monitoring COVID-19 infection patterns.

The research team analyzed wastewater data obtained from the city of Sapporo in northern Japan between April 2021 and September 2023. The dataset featured high sensitivity (100 times greater than the standard method) and high reproducibility (standard deviation below 0.4 at log10 values) and was supported by a substantial sample size of 15 samples per week, totaling 1,830 samples over a sufficient survey period of two and a half years. The correlation coefficient between the number of infected individuals and the viral concentration in the wastewater was 0.87, indicating that this method effectively tracks regional infection trends. Additionally, the research team concluded desirable survey frequency requires at least three samples, preferably five samples, per week.

The study provides detailed guidance on wastewater surveillance methodologies for understanding infection trends, focusing on data processing, analytical sensitivity, and survey frequency. As wastewater surveillance during normal times becomes more widely implemented and its results increasingly published, this study's findings are expected to serve as valuable resources for decision making.

Source: Osaka University

Journal reference: Murakami, M., et al. (2024) Evaluating survey techniques in wastewater-based epidemiology for accurate COVID-19 incidence estimation. The Science of the Total Environment. doi.org/10.1016/j.scitotenv.2024.176702. www.sciencedirect.com/science/article/pii/S0048969724068591?via%3Dihub

[elbiotipo]

fwiw and i have no idea what the artists are doing with it, a lot of the libraries that researchers are currently using to develop deep learning models from scratch are all open source built upon python, i'm sure monsanto has its own proprietary models hand crafted to make life as shitty as possible in the name of profit, but for research there's a lot of available resources library and dataset wise in related fields. It's not my area per se but i've learnt enough to get by in potentially applying it to my field within science, and largely the bottleneck in research is that the servers and graphics cards you need to train your models at a reasonable pace are of a size you can usually only get from google or amazon or facebook (although some rich asshole private universities from the US can actually afford the cost of the kind of server you need. But that's a different issue wrt resource availability in research in the global south. Basically: mas plata para la universidad pública la re puta que los parió)

Yes, one great thing about software development is that for every commercially closed thing there are open source versions that do better.

The possibilities for science are enormous. Gigantic. Much of modern science is based on handling huge amounts of data no human can process at once. Specially trained models can be key to things such as complex genetics, especially simulating proteomes. They already have been used there to incredible effect, but custom models are hard to make, I think AIs that can be reconfigured to particular cases might change things in a lot of fields forever.

I am concerned, however, of the overconsumption of electronics this might lead to when everyone wants their pet ChatGPT on their PC, but this isn't a thing that started with AI, electronic waste and planned obsolescence is already wasting countless resources in chips just to feed fashion items like iphones, this is a matter of consumption and making computers be more modular and longer lasting as the tools they are. I've also read that models recently developed in China consume much, much less resources and could potentially be available in common desktop computers, things might change as quickly as in 2 years.

[probablyasocialecologist]

The World Food Summit of 1996 approached food security through the principles of ensuring there is enough safe and nutritious food that can be accessed daily to meet healthy dietary needs and food preferences. By definition, this is a desirable and worthy goal. However, in the years since, food security has developed into a paradigm which does not question the underlying power dynamics and the reproduction of material conditions that make food insecurity a permanent feature of the global order. At its core, the food security paradigm deals only with access to food, without challenging the political and economic structures that determine and control access, as well as distribution.  By failing to address the root causes of hunger and famine, the food security paradigm makes it impossible to end hunger globally. Of course, many people worldwide possess food security, but this is restricted to increasingly limited geographic pockets. In terms of the people localised in one area, food vulnerability is influenced and determined by class, race, gender and, of course, citizenship status. Globally, “underdevelopment” and “de-development” lead to widespread food insecurity across areas. Another problem with the food security paradigm is that it is easily co-opted to generate partial answers that pose no threat to the corporate food system, or worse, that even open up new profit opportunities. Accelerated by other crises, the food security paradigm becomes ever more dependent on aid, be it through direct food delivery, cash transfers or small development projects that cannot compete with the food giants and their price-setting powers. In practice, a “science of food security” emerges, one which takes as its focus calories and the output that is compatible with precision agriculture having the aim to increase crop yields and to assist management decisions using high technology sensor and analysis tools. This model tends to be reliant on “Green Revolution” technologies that rely on chemical fertilisers and pesticides and that are tied to colonial projects and corporations, in order to optimise resources in aid response and/or development projects.  In this rationale, food insecurity can be addressed by reaching optimum yields of certain crops that should meet the demand for fats, fibres and protein. All of this is carefully managed and data-driven. Precision farming is advocated by the Alliance for a Green Revolution in Africa (AGRA) with the objective of optimising, “agricultural value chains […] critical in advancing food and nutrition sufficiency without increasing the size of land under cultivation.” The framing of food that reduces it only to “optimal input” relegates vital elements of food production and the culture of eating, like territory ownership, taste, heritage, care, well-being and connection as secondary. This reductionist approach has, though, proved useful to corporate agriculture, since it reinforces the case for genetically modified crops (GMOs), more efficient fertilisers, and the standardisation of food production for market purposes. Advocates of plant breeding technologies (including GMOs and hybrid seeds) argue that government overregulation is an obstacle to achieving food security. Overregulation, as the argument goes, denies populations the opportunity to grow crops that have increased nutrient use efficiency and are more resilient to climate shocks. 

[...]

The paradigm of food security is about optimising productivity. It’s true that productivity matters – after all, feeding the world requires enormous quantities of food. But if productivity is approached solely as a technological problem, it reinforces the tendency to fragment the quantitative and qualitative aspects of food production and consumption. On the quantitative side, production for food security is viewed as a challenge of multiplication. Whereas division, that is, distribution of food, is left to logistical planning. This ignores what Raj Patel identified in his influential 2007 book Stuffed and Starved, as the bottleneck of power that concentrates international food distribution among a small set of corporations. This bottleneck excludes the poor and small-scale food producers from decision-making. It also normalises worrying tendencies, such as an overreliance on industrial animal exploitation as a protein source, which has direct health implications, as well as longer term consequences like the proliferation of new viruses, greenhouse gas emissions and inefficient use of water and soil.

28 May 2024

[structuredsucc]

Let's talk about that autism registry thing...

I've seen a few posts now where people are talking about this RFK autism registry thing in very all-or-nothing ways, to the point of wondering if their life is over.

I'm not going to pull punches, it's probably going to be harmful, but your fight; our fight isn't over.

Let's talk about it

The context:

RFK has made a promise to find out the cause of increasing autism rates by September. To do this, he isn't asking experts, he wants to compile a database of autistic people and our private health records to study the issue.

RFK would like to make this database available to researchers (including researchers he hand selects) to study increasing autism rates.

I'm not going to lie. This is bad. This is bad science. This is probably going to be harmful to autistic people and community… but it's also complicated.

The complications:

Having a disease database isn't a new thing, even in the US. There are databases like this for more threatening conditions to public health, such as cancers, infectious diseases, and bio-threats already. They are used in science and public health monitoring and can be good.

In a number of countries (especially Scandinavia), public health databases like these exist, including of autistic, ADHD, trans, and other vulnerable groups. These databases can do a lot to further research with vulnerable communities…

but they can also be a massive liability.

The Difference:

The current US administration.

The current US administration has clearly and repeatedly shown they don't mind (or actively wish to) harming vulnerable communities. RFK has stated numerous times his dangerous and dangerously misinformed opinions about autistic people.

So, while public health databases can be used for good things, I don't think this one is going to be. However, it probably isn't the end times either.

The Probable Future:

More than likely this database is going to be used (first) to further disinformation about autism, including vaccine myths. Spreading disinformation about autism, especially from a government, is going to be immeasurably harmful, no doubt.

But we've been here before. We've fought mis/disinformation and we have the tools and community to do it again.

What worries people:

What worries people is what happens after. The current US administration has been shown repeatedly that they are willing to use private data that they have on citizens to harm people.

They have used data like this for ICE raids and targeted doxing already. No one can promise that this database will not be used the same way.

That said, the path between where we are now and using this proposed database for targeted harm is windy and uncertain. We are not there yet, and there are still things that we can do.

What we can do:

1) Take care of yourself.

Make sure that you are meeting your physical, mental, and sensory needs, at least to a bare minimum level, before giving too many spoons to others.

2) Build community.

You probably have autistic family members. Talk to them.

Is there an autistic-led organization in your area? Autistic-led peer support? Reach out to them. Talk to other autistic people and share resources

If there isn't autistic-led organizations in your area, can you start something? It doesn't have to be big or perfect or consistent.

Heck, even an autistic-run D&D campaign, or social group, or bowling night would build community, just so long as it gets people meeting!

Just don't leave anyone behind. Include the most vulnerable (those in group/care homes, non-speaking, etc.) among us in these communities.

3) Communicate with your physical or mental health care team.

They may not be able to amend past health records, but they can certainly change the notes they take in future.

Be careful though, as these notes are an important part of disability programs and may be necessary for those purposes.

4) Continue making noise.

Get the media involved. Be loud at people who could be involved with the database (researchers, government officials, committee members). Talk to your representatives.

5) What'd I forget?

What are other things we can do the push back in this moment?

[mariacallous]

With U.S. President Donald Trump, many high-tech titans have decided that now—after their coffers overflowing—Americans don’t need much government. Leading the charge to dismantle it is Elon Musk. His role is especially jarring because Silicon Valley was built on the government’s largesse. A booming high-tech sector—one of the signature achievements of the modern economy—wouldn’t have happened without the administrative state that Trump is seeking to root out.

The history of Silicon Valley exposes the grave dangers posed by the war on government. The hazard is that as a result of this push, Trump succeeds in breaking apart the marriage between Washington and the technology industry that has helped make America great.

The road to high tech really started to be built during World War II. In 1945, Vannevar Bush, who had directed the U.S. Office of Scientific Research and Development during the war, captured the zeitgeist of the era when he published “Science: The Endless Frontier,” which offered a declaration of principle for the government supporting scientific education. The report, submitted to President Harry Truman, explained why government support for research was so important to national security and the economic well-being of the nation. “The pioneer spirit is still vigorous within this nation,” Bush wrote in the letter that accompanied the report. “Science offers a largely unexplored hinterland for the pioneer who has the tools for his task. The rewards of such exploration both for the Nation and the individual are great. Scientific progress is one essential key to our security as a nation, to our better health, to more jobs, to a higher standard of living, and to our cultural progress.”

Much of the development of large mainframe computing systems was born of defense needs. While mainframe systems were being built in the early 1930s, during the war, the U.S. Army and several other defense units developed the Electronic Numerical Integrator and Computer (ENIAC) under the direction of Maj. Gen. Gladeon Barnes. Congress devoted massive resources (today’s equivalent of millions in current) dollars to the construction of what would become the first general-use computer. The most important initial function of ENIAC, which was completed in 1946 by University of Pennsylvania scholars John Mauchly and J. Presper Eckert, was its ability to provide cutting-edge calculations about the trajectories of weapons. Before the project ended, the government discovered ways to use ENIAC for a wide range of jobs, including advanced weather prediction and wind tunnel design. With funding from the Census Bureau, Mauchly and Eckert next worked on the Universal Automatic Computer (UNIVAC), resulting in a digital computer allowing for data processing and storage methods that were new and extremely beneficial to industry. With CBS anchor Walter Cronkite standing by, UNIVAC, which weighed a whopping 16,000 pounds, famously predicted early on election evening in 1952 that Dwight Eisenhower would defeat Adlai Stevenson by a landslide. A computer star was born. The machine would even appear on the cover of a Superman comic book.

Throughout the early Cold War in the 1940s and 1950s, the federal government poured resources into the production of knowledge. The GI Bill of Rights (1944) vastly expanded the student body by covering the cost of enrollment and more for veterans, many of whom were first-generation students. In 1950, Truman signed legislation creating the National Science Foundation, an institution that complemented the National Institutes of Health by aiding nonmedical science and engineering. Their shared mission was to “promote the progress of science; to advance the national health, prosperity and welfare; and to secure the national defense.” Eisenhower, a Republican, worked with congressional Democrats such as Sen. Lyndon B. Johnson to respond to the Soviet Union’s successful launch of the Sputnik satellite in 1957 by building on this precedent. The National Defense Education Act (1958) financed student loans, graduate fellowships, and research funds. By the early 1960s, with substantial help from the government, U.S. universities were booming and considered to be among the finest institutions of learning anywhere in the world. As the Cold War kept heating up, one area where Americans were clearly ahead was on the campus.

Without the government-industry connection that emerged from this era, there would be no internet. While there may still be people debating whether former Vice President Al Gore invented the internet, there is no dispute that the federal government did. The Defense Advanced Research Projects Agency (DARPA), established in 1958, undertook high-risk, large-scale research, cooperating with private firms, that had the potential to produce enormous payoffs. DARPA was central to the Advanced Research Projects Agency Network (ARPANET) in the late 1960s, which constituted the first advanced computer network. Much of the drive for the military had been the desire for a functional network that could survive a nuclear attack. ARPANET was the basis for the modern internet. The National Science Foundation announced a distinct section, called NSFNET, in 1986. The foundation connected five supercomputer centers and granted academic network’s access. The project was considered to have been the “backbone” for the creation of the commercial internet. Other notable computer innovations also grew out of this operation. DARPA dollars facilitated the Stanford Research Institute’s making of the mouse, a technology that made it easier for an individual without great technical expertise to interface with computers. In 1991, Congress passed the High-Performance Computing Act—legislation that Gore helped move—which funded a team of programmers at the University of Illinois’s National Center for Supercomputing Applications that helped vastly expand the internet. Marc Andreessen, one of the engineers who co-created Mosaic and Netscape, acknowledged in 2000, “If it had been left to private industry, it wouldn’t have happened, at least, not until years later.”

Indeed, Silicon Valley would not have become what it is today without the government. The DARPA-Stanford research partnership, as the historian Margaret O’Mara has brilliantly recounted in Cities of Knowledge and The Code, is a big reason why the university emerged as such a powerhouse in high-tech education and research. Government money fueled the transformation of a formerly sleepy region, which O’Mara reminds us would have once been improbable to imagine as a hub of big inventions and money. A series of Stanford leaders, including provost Frederick Terman, opened their arms to the federal coffers and shepherded the Stanford Research Park into its current incarnation.

Not only was Stanford built up with government monies, but many of the companies that have littered the landscape in northern California had Washington to thank. Fairchild Semiconductor, established in San Jose in 1957, took form with Air Force and NASA contracts. NASA’s ongoing investment in the integrated circuits that it and other companies produced allowed costs to become accessible and for the semiconductor industry to emerge. Federal dollars during the 1980s and 1990s that were tied to programs such as President Ronald Reagan’s Strategic Defense Initiative—a massive laser missile shield that would protect the United States from nuclear attack, which critics derided as “Star Wars”—resulted in all sorts of computer innovations not envisioned by the administration’s plan. Though stories about Steve Jobs and Steve Wozniak working out of a garage capture our entrepreneurial imaginations, the role of the administrative state continues to loom large over the entire region. “From the marble halls of Washington and the concrete canyons of Wall Street,” O’Mara writes in The Code, Silicon Valley was made by many hands. Other “cities of knowledge,” including Cambridge, Massachusetts; Philadelphia; and Atlanta, were similar beneficiaries of government.

The federal government has helped high tech in many other ways besides policies directly related to computers and the internet. Immigration reforms, for instance, that opened the doors to high-skilled foreign-born immigrants resulted in the arrival of people who helped build the computing products that the entire world depends on today. The Hebrew Immigrant Aid Society helped a young Sergey Brin and his family obtain a visa to emigrate from the Soviet Union in 1979. With that, Google was born. Musk was able to finish his education at the University of Pennsylvania with a student visa and stay in the United States because of an H1-B visa. Yahoo co-founder Jerry Yang immigrated with his family from Taiwan in 1978. The Small Business Investment Incentive Act (1980) provided valuable dollars to Silicon Valley firms as they struggled to make a name for themselves.

Indeed, Musk’s company Tesla benefited from government assistance. In 2009, a critical moment for the company, Tesla received $465 million in low-interest loans from the U.S. Energy Department that it used to construct the Model S. Electric vehicle tax credits have grown consumer demand for his and other vehicles. Federal research grants played a role in the different components that make up these cars.

The federal government and the high-tech industry have stood side by side for decades. And the high-tech story has happened many times over, often in some of what have become the country’s most conservative areas. In From Cotton Belt to Sunbelt, historian Bruce J. Schulman traces how the revitalization of the South and Southwest, ground zero for the modern conservative movement of the post-1960s era, was built on defense contracts and military bases. Reagan’s presidency, which pushed politics rightward, derived electoral profits from massive congressional investments made over the decades after the war.

While many agree on the importance of markets, the hand of government—sometimes hidden from view—has been equally essential to economic success. The history of high tech has revolved around a genuine partnership between markets and government, not one or the other. To destroy the partnership threatens to destroy what has made the U.S. economy great. Every American will be forced to pay the cost.

[sailorsoons]

what are "other forms of Ai" that's bad? isn't it just generative Ai that's the issue? everything u linked is about generative Ai and not "other forms of Ai". /gen

Hi! Great question and you're right - I only linked some of the resources on Generative AI. Additional forms of AI are thins like predictive AI, machine learning AI, and general AI that performs tasks like web scraping, automated data collection, crowdsourcing, and chatbots/AI assistants. All of these different types of AI have the same issue as generative AI in terms of environmental damages.

One could argue that everything we do causes environmental damage because it does, but at the current rate that countries, companies, governments and individual people use the different AI tools listed above, we are severely impacting energy, water and metal resources. The amount of power and sheer materials used to create the servers and CPUs that output all of the above tools is insane.

Here are some additional resources on cases where different types of AI have had negative outcomes:

Predictive AI Use in Law Enforcement Agencies

AI Screenings make it harder to get hired

AI Screenings make it harder to get hired part II

Machine learning/predictive AI is bad for science

This article covers a ton of different positives and negatives of AI

In general, there are a lot of potentials and futures where AI could make a positive impact, but at a global scale, it's not being used that way. The last article linked offers some ways in which AI could actually be useful, especially in the medical field, but at large, AI is disproportionately used with negative results and at a high cost.