#AI and Data Science Projects

[technology-123s-blog]

Deep Learning Projects for Students - Takeoff Projects

Deep Learning Projects are exciting and advanced applications of artificial intelligence that solve complex problems by mimicking the way humans think. At Takeoff Projects, we provide a platform for students and professionals to explore and work on innovative deep learning projects that are both educational and practical. These projects involve training neural networks to analyze large amounts of data and make intelligent decisions.

Some popular deep learning projects include image recognition, where models identify objects or faces in pictures, and natural language processing, which helps in building chatbots or translating languages. Deep learning is also used in healthcare to analyze medical images like X-rays to detect diseases, and in self-driving cars to recognize objects on the road and ensure safe navigation.

At Takeoff Projects, we guide learners through real-world projects such as creating speech recognition systems, building recommendation engines like those used by Netflix or Amazon, and designing AI models for time-series forecasting like stock price prediction. We simplify these concepts with hands-on support, making them easy to understand and implement.

We also focus on innovative areas like Generative Adversarial Networks (GANs), which can create realistic images or enhance low-resolution photos, and robotics, where deep learning enables machines to perform tasks like sorting or assembly. These projects not only build technical skills but also prepare learners for a bright future in AI and data science.

Whether you are a beginner or an advanced learner, Takeoff Projects helps you take the first step toward mastering deep learning. By working on these projects, you can gain practical experience and showcase your expertise, opening up exciting career opportunities in this rapidly growing field. Let’s take off into the world of deep learning Projects together!

[bigleapblog]

Your Guide to B.Tech in Computer Science & Engineering Colleges

In today's technology-driven world, pursuing a B.Tech in Computer Science and Engineering (CSE) has become a popular choice among students aspiring for a bright future. The demand for skilled professionals in areas like Artificial Intelligence, Machine Learning, Data Science, and Cloud Computing has made computer science engineering colleges crucial in shaping tomorrow's innovators. Saraswati College of Engineering (SCOE), a leader in engineering education, provides students with a perfect platform to build a successful career in this evolving field.

Whether you're passionate about coding, software development, or the latest advancements in AI, pursuing a B.Tech in Computer Science and Engineering at SCOE can open doors to endless opportunities.

Why Choose B.Tech in Computer Science and Engineering?

Choosing a B.Tech in Computer Science and Engineering isn't just about learning to code; it's about mastering problem-solving, logical thinking, and the ability to work with cutting-edge technologies. The course offers a robust foundation that combines theoretical knowledge with practical skills, enabling students to excel in the tech industry.

At SCOE, the computer science engineering courses are designed to meet industry standards and keep up with the rapidly evolving tech landscape. With its AICTE Approved, NAAC Accredited With Grade-"A+" credentials, the college provides quality education in a nurturing environment. SCOE's curriculum goes beyond textbooks, focusing on hands-on learning through projects, labs, workshops, and internships. This approach ensures that students graduate not only with a degree but with the skills needed to thrive in their careers.

The Role of Computer Science Engineering Colleges in Career Development

The role of computer science engineering colleges like SCOE is not limited to classroom teaching. These institutions play a crucial role in shaping students' futures by providing the necessary infrastructure, faculty expertise, and placement opportunities. SCOE, established in 2004, is recognized as one of the top engineering colleges in Navi Mumbai. It boasts a strong placement record, with companies like Goldman Sachs, Cisco, and Microsoft offering lucrative job opportunities to its graduates.

The computer science engineering courses at SCOE are structured to provide a blend of technical and soft skills. From the basics of computer programming to advanced topics like Artificial Intelligence and Data Science, students at SCOE are trained to be industry-ready. The faculty at SCOE comprises experienced professionals who not only impart theoretical knowledge but also mentor students for real-world challenges.

Highlights of the B.Tech in Computer Science and Engineering Program at SCOE

Comprehensive Curriculum: The B.Tech in Computer Science and Engineering program at SCOE covers all major areas, including programming languages, algorithms, data structures, computer networks, operating systems, AI, and Machine Learning. This ensures that students receive a well-rounded education, preparing them for various roles in the tech industry.

Industry-Relevant Learning: SCOE’s focus is on creating professionals who can immediately contribute to the tech industry. The college regularly collaborates with industry leaders to update its curriculum, ensuring students learn the latest technologies and trends in computer science engineering.

State-of-the-Art Infrastructure: SCOE is equipped with modern laboratories, computer centers, and research facilities, providing students with the tools they need to gain practical experience. The institution’s infrastructure fosters innovation, helping students work on cutting-edge projects and ideas during their B.Tech in Computer Science and Engineering.

Practical Exposure: One of the key benefits of studying at SCOE is the emphasis on practical learning. Students participate in hands-on projects, internships, and industry visits, giving them real-world exposure to how technology is applied in various sectors.

Placement Support: SCOE has a dedicated placement cell that works tirelessly to ensure students secure internships and job offers from top companies. The B.Tech in Computer Science and Engineering program boasts a strong placement record, with top tech companies visiting the campus every year. The highest on-campus placement offer for the academic year 2022-23 was an impressive 22 LPA from Goldman Sachs, reflecting the college’s commitment to student success.

Personal Growth: Beyond academics, SCOE encourages students to participate in extracurricular activities, coding competitions, and tech fests. These activities enhance their learning experience, promote teamwork, and help students build a well-rounded personality that is essential in today’s competitive job market.

What Makes SCOE Stand Out?

With so many computer science engineering colleges to choose from, why should you consider SCOE for your B.Tech in Computer Science and Engineering? Here are a few factors that make SCOE a top choice for students:

Experienced Faculty: SCOE prides itself on having a team of highly qualified and experienced faculty members. The faculty’s approach to teaching is both theoretical and practical, ensuring students are equipped to tackle real-world challenges.

Strong Industry Connections: The college maintains strong relationships with leading tech companies, ensuring that students have access to internship opportunities and campus recruitment drives. This gives SCOE graduates a competitive edge in the job market.

Holistic Development: SCOE believes in the holistic development of students. In addition to academic learning, the college offers opportunities for personal growth through various student clubs, sports activities, and cultural events.

Supportive Learning Environment: SCOE provides a nurturing environment where students can focus on their academic and personal growth. The campus is equipped with modern facilities, including spacious classrooms, labs, a library, and a recreation center.

Career Opportunities After B.Tech in Computer Science and Engineering from SCOE

Graduates with a B.Tech in Computer Science and Engineering from SCOE are well-prepared to take on various roles in the tech industry. Some of the most common career paths for CSE graduates include:

Software Engineer: Developing software applications, web development, and mobile app development are some of the key responsibilities of software engineers. This role requires strong programming skills and a deep understanding of software design.

Data Scientist: With the rise of big data, data scientists are in high demand. CSE graduates with knowledge of data science can work on data analysis, machine learning models, and predictive analytics.

AI Engineer: Artificial Intelligence is revolutionizing various industries, and AI engineers are at the forefront of this change. SCOE’s curriculum includes AI and Machine Learning, preparing students for roles in this cutting-edge field.

System Administrator: Maintaining and managing computer systems and networks is a crucial role in any organization. CSE graduates can work as system administrators, ensuring the smooth functioning of IT infrastructure.

Cybersecurity Specialist: With the growing threat of cyberattacks, cybersecurity specialists are essential in protecting an organization’s digital assets. CSE graduates can pursue careers in cybersecurity, safeguarding sensitive information from hackers.

Conclusion: Why B.Tech in Computer Science and Engineering at SCOE is the Right Choice

Choosing the right college is crucial for a successful career in B.Tech in Computer Science and Engineering. Saraswati College of Engineering (SCOE) stands out as one of the best computer science engineering colleges in Navi Mumbai. With its industry-aligned curriculum, state-of-the-art infrastructure, and excellent placement record, SCOE offers students the perfect environment to build a successful career in computer science.

Whether you're interested in AI, data science, software development, or any other field in computer science, SCOE provides the knowledge, skills, and opportunities you need to succeed. With a strong focus on hands-on learning and personal growth, SCOE ensures that students graduate not only as engineers but as professionals ready to take on the challenges of the tech world.

If you're ready to embark on an exciting journey in the world of technology, consider pursuing your B.Tech in Computer Science and Engineering at SCOE—a college where your future takes shape.

[uditprajapati7685]

Pickl.AI offers a comprehensive approach to data science education through real-world case studies and practical projects. By working on industry-specific challenges, learners gain exposure to how data analysis, machine learning, and artificial intelligence are applied to solve business problems. The hands-on learning approach helps build technical expertise while developing critical thinking and problem-solving abilities. Pickl.AI’s programs are designed to prepare individuals for successful careers in the evolving data-driven job market, providing both theoretical knowledge and valuable project experience.

[delicatelysublimeforester]

The City Nature Challenge: Embracing Earth’s Wonders, One Observation at a Time

[ejobindia-blog]

AI & ML Training with Live Projects in Kolkata – Ejobindia

Ejobindia's AI & ML training program is tailored for both beginners and professionals aiming to delve into the world of AI. The course emphasizes hands-on learning, ensuring that students not only grasp theoretical concepts but also apply them in real-world scenarios.​

Course Highlights

Duration: The program spans 100 hours, providing an in-depth understanding of AI and ML concepts.​

Course Fee: The total fee for the course is ₹30,500.​

Curriculum Includes:

Fundamentals of AI & ML

Prompt Engineering

Large Language Models (LLMs)

Industry Use Cases

Vector Databases

Hands-on assignments and live projects​

This structured approach ensures that students gain both the theoretical knowledge and practical skills required in the AI industry.​Home

Why Choose Ejobindia?

Industry-Relevant Training: The curriculum is designed in collaboration with industry experts to ensure relevance in today's job market.​Home

Experienced Trainers: Learn from professionals with extensive experience in AI and ML.​

Placement Support: Ejobindia boasts partnerships with over 100 hiring companies and facilitates approximately 300 placements annually.​eJobIndia+1eJobIndia+1

Live Projects: Gain practical experience by working on real-world projects, enhancing your portfolio and confidence.​

Flexible Learning Modes: Choose between online and offline classes based on your convenience.​eJobIndia

Upcoming Batches

Ejobindia regularly updates its batch schedules. For the most recent information on upcoming batches, it's recommended to visit their official website or contact them directly.​

How to Enroll

To enroll in the AI & ML Training with Live Projects in Kolkata at Ejobindia:

Visit the Official Website: Navigate to Ejobindia's AI & ML Training Page.​eJobIndia

Contact: For direct inquiries, you can call them at 9830228812 / 9830125644 or email via the contact form on their website.​eJobIndia

Fill Out the Enrollment Form: Provide the necessary details and choose your preferred batch timing.​

Embark on your AI journey with Ejobindia and equip yourself with the skills to thrive in the ever-evolving tech landscape.

[anonymousdormhacks]

Google says alexander "cheated on his wife" hamilton rights ig

[ambrosiaventures]

How Pharmaceutical Consulting Can Help Launch Your New Product Successfully

Ambrosia Ventures, we ensure your product launch achieves maximum impact by utilizing our expertise in biopharma consulting, which makes us a trusted pharmaceutical consulting service provider in the US. Here's the way to transform your product launch strategy into a blueprint for success through pharmaceutical consulting services

[nnctales]

AI Consulting Business in Construction: Transforming the Industry

The construction industry is experiencing a profound transformation due to the integration of artificial intelligence (AI). The AI consulting business is at the forefront of this change, guiding construction firms in optimizing operations, enhancing safety, and improving project outcomes. This article explores various applications of AI in construction, supported by examples and statistics that…

[cymetrixsoftware]

CHOOSE THE RIGHT STAFF AUGMENTATION MODEL FOR YOUR DATA ANALYTICS PROJECTS

Are you tired of struggling to find the right talent for your Data Analytics project? You're not alone. The high demand for analytical skills and the shortage of qualified candidates make bridging the skill gap challenging. However, the right staff augmentation model can be the solution you need. At Cymetrix, we specialize in providing tailored staff augmentation services for Data Science, Analytics, Salesforce, and AI. An approach that allows you to seamlessly integrate highly skilled professionals into your team, ensuring you have the expertise required for project success. Whether you need project-based, skill-based, capacity-based, strategic, or hybrid augmentation, we've got you covered. With Cymetrix, you stay in control while benefiting from external talent, fostering a cohesive unit focused on your goals. Find out how our flexible solutions can help you build the perfect team and ensure the success of your Data Analytics projects. Keep reading to learn more about choosing the right staff augmentation model for your data analytics projects.

[oneofthosecrazycatladies]

Okay a couple weeks ago I started this post trying to keep track of all the stuff going on in order to help remind us of everything that’s happened when the next election comes around. Well, because there’s just so much going on, I’ve realized trying to cram it all into one post isn’t going to work. So I’m going to do a new post every month and include links to the previous ones.

So here goes…

January 2025

February 2025

Donald Trump has enforced his tariffs on Mexico, Canada, and China. [x]

Donald Trump has put Mexico tariffs on hold for one month. [x]

Donald Trump allowed Elon Musk to begin dismantling USAID. [x]

Congress is voluntarily giving up its power and allowing Trump to make unilateral decisions. [x]

Darren Beattie has been made Under Secretary of State. [x]

Everything that Donald Trump has done so far lines up with Project 2025 [x]

The White House is drafting an executive order to eliminate the Department of Education [x]

Elon Musk, who nobody voted for or elected, has, essentially, hacked the government. [x]

El Salvador has agreed to take US deportees of any nationality. [x]

US Representative Andy Biggs is proposing a bill to abolish OSHA. [x]

Pam Bondi has been confirmed as Attorney General [x]

Donald Trump doesn’t think Palestinians should return to Gaza. [x]

Donald Trump says he’ll use US troops to “take over” the Gaza Strip. [x]

A federal judge has blocked Donald Trump’s executive order to end birthright citizenship. [x]

Donald Trump has banned trans women from women’s sports [x]

Donald Trump sanctions the International Criminsl Court. [x]

A judge has paused the federal “buyouts” [x]

DOGE: Member of DOGE resigns [x]

DOGE has been given access to the Department of Energy. [x]

Miscellaneous news about Elon Musk [x]

DOGE is using AI to infiltrate the Department of Education [x]

Russell Vought, author of Project 2025, has been confirmed as Director of OMB [x]

Democrats in Congress have introduced the Taxpayer Data Protection Act [x]

Donald Trump has flagged the words “women” “diverse” and “historically” from studies done by the National Science Foundation. [x]

New Mexico Representative Melanie Stansbury has introduced the Nobody Elected Elon Musk Act [x]

Democratic Congressional leaders have introduced the Stop the Steal Act [x]

Donald Trump has called for a review of funding for the United Nations [x]

Federal agencies are barred from celebrating Black History Month [x]

Donald Trump has frozen aid to South Africa and accused the South African government of racism against white South Africans [x]

Donald Trump wants to use Leavenworth Prison as a migrant detention facility and have it run by a for-profit company known for its numerous human rights violations. [x] [x]

Trump has told the Treasury to stop making pennies. [x]

Representative Mark Pocan (D-WI) proposes the E.L.O.N. M.U.S.K. Act (which stands for Eliminate Looting of Our Nation by Mitigating Unethical State Kleptocracy) [x]

Employees of the Consumer Financial Protection Bureau were told to stop all work and are now being told to stay home. [x]

Trump will impose 25% tariffs on steel and aluminum. [x]

Trump says Palestinians won’t be allowed back in Gaza if the US takes it over [x]

Tulsi Gabbard has been confirmed as director of national intelligence. [x]

Representative Buddy Carter (R-GA) has proposed a bill to change the name of Greenland to Red, White & Blue Land [x]

The DOJ has dropped the corruption charges against New York City mayor Eric Adams. [x]

An AP News reporter has been banned from the White House for using Gulf of Mexico instead of Gulf of America in its reporting. [x][x]

Senators Deb Fischer (R-NE) and Angus King (I-ME) are pushing for a tax credit that would encourage businesses to offer paid family leave. [x]

Representative Sara Jacobs (D-CA) has introduced the Protect US National Security Act [x]

The State Department (taxpayers) is paying Elon Musk $400 million for cybertrucks. [x]

Robert F. Kennedy Jr. has been confirmed as HHS Secretary. [x]

Trump is conducting a mass firing of the federal workforce. [x]

Senator Ted Cruz (R-TX) is creating a list of all the ‘woke’ science he wants to get rid of. [x]

References to transgender have been removed from the Stonewall National Monument. [x]

A 71 year old refugee living in Thailand has died because of the USAID freeze. [x][x]

Trump’s proposed tax cuts will add trillions to US debt. [x]

Trump is defying the court order to reopen USAID. [x]

Trump has stopped the CDC’s flu vaccine campaign. [x]

Trump is suing Brazil’s Supreme Court because of Brazil’s battles with Elon Musk over Twitter/X. [x]

Kash Patel has been confirmed as FBI director. [x]

Trump orders FEMA to stop their work with making homes better at withstanding natural disasters. [x]

Kash Patel will be named chief of the ATF [x]

Trump has tried to make independent agencies no longer independent [x]

$200 million of taxpayer money was used on a pro-Trump anti-migrant ad [x]

The House of Representstives passed a bill that gives more than $4 trillion in tax cuts for the wealthy and cuts the budget for Medicaid by 80% [x]

Here’s a summary of Trump’s executive orders so far [x]

The Trump administration has issued travel bans for trans athletes [x]

Trump administration is telling federal agencies to prepare for more mass layoffs [x]

Elon Musk joined Trump’s first cabinet meeting. [x]

Trump is offering “gold cards” to wealthy foreigners [x]

Kash Patel names Dan Bongino as Deputy Director of the FBI. [x]

Senator Mike Lee (R-UT) has proposed legislation for the US to leave the United Nations [x]

Judge rules mass firings of federal workers is unlawful [x]

The Pentagon orders all transgender people to be removed from the military [x]

Representative Victoria Spartz (R-IN) was going to vote against the budget bill that would cut nearly $1 trillion from Medicaid; then she got a phone call from Trump who apparently screamed at and threatened her; she then voted yes on the bill [x]

Trump administration has cancelled boot camps for women training to become Wildland firefighters [x]

Here’s a link to the Project 2025 Policy Agenda that Donald Trump claimed he didn’t know anything about.*

*He only claimed he didn’t know anything about it after it proved to be deeply unpopular with the general public.

I’m also including directories for both the House of Representatives and the Senate. That way, if you’re so inclined, you can also track the individual actions of every Senator and Representative.

Miscellaneous News

Representative Nancy Mace (R-SC) repeatedly uses a transphobic slur on the Congressional floor. [x]

Clarence Thomas is…being Clarence Thomas *sigh* [x]

Donald Trump fired the Chair of the Kennedy Center and named himself as the new Chair [x]

Trump said that no group of people in the history of America has been treated worse than the way the January 6th insurrectionists have been treated. [x]

Some people are impersonating ICE agents and harassing & assaulting people of color [x][x]

Trump’s mass deportation is hitting a wall [x]

The Trump administration’s incompetence is coming back to bite them. [x]

Target has been facing backlash for rolling back its DEI initiatives. [x]

Donald Trump Has Already Spent $10.7 Million Of Taxpayer Money Playing Golf [x]

The Kennedy Center cancelled a performance of the Gay Men’s Chorus of Washington DC [x]

21 DOGE employees have resigned [x]

Musk’s new conflict of interest [x]

Trump posted an AI-created video about his plans for Gaza [x]

Here’s a Washington Post story about the migrants sent to Guantanamo Bay and the conditions they’re facing [x]

Trump supporters are calling for “processing camps” and private militias to go after migrants. [x]

Representative Cory Mills (R-FL) has been accused of assault and the Department of Justice is refusing to investigate [x]

A child has died in the measles outbreak in Texas [x]

China and Russia are trying to recruit disgruntled federal employees [x]

Elon Musk is trying to force the FAA to get rid of their contract with Verizon in favor of a contract with his company, Starlink [x]

Elon Musk makes $38 billion in government contracts [x]

Trump thinks that Andrew Tate is a totally okay guy [x]

The director of the Defense Health Agency abruptly retired [x]

March-June 2025

Once again, please feel free to let me know about anything I’ve missed. With this era of constant news we live in, it can be easy to forget so let’s give our future selves a little help!

[vlruso]

Youve Hit a Wall in Your Data Project Now What?

📢 [New Blog Post] "You've Hit a Wall in Your Data Project, Now What?" Dealing with obstacles in data analytics development can be challenging. Unexpected outcomes, abnormal record numbers, or incorrect results can leave you scratching your head. So, how do you tackle these problems when there are no clear indications? Drawing from my 6+ years of experience in analytics roles, I've gathered some effective strategies to overcome these obstacles. In the latest blog post, I'll share successful techniques that will make you more savvy in dealing with such situations. Take a moment to step away from the problem and gain a fresh perspective. Sometimes, a short walk can help you see things from a different angle. Remember, maintaining a calm and open-minded approach is key when facing data project hurdles. Read the full article here to discover the strategies, tips, and tricks for overcoming obstacles in your data projects: [Link to the article](https://ift.tt/hgSxfEV) #dataanalytics #dataprojects #problemsoving #analyticsdevelopment List of Useful Links: AI Scrum Bot - ask about AI scrum and agile Our Telegram @itinai Twitter -  @itinaicom

[aboutaiart]

Ai Photos and Information.

[sunarryn]

Dp X Marvel #6

They called him Wraith.

Not Phantom. Not Fenton. Not Danny. Those names belonged to a ghost of a boy that never made it out of a cold, steel lab buried beneath the earth—forgotten by the world, forsaken by the stars. Wraith was something else. A project. A weapon. An experiment that should have failed but didn’t. The product of every nightmare HYDRA ever dared to dream. Not even the Red Room could engineer something so devastating. Not even Arnim Zola’s data-crazed AI mind could fathom the scope of him. Even the Winter Soldier—their perfect killer—trembled at the mere scent of Wraith in the air. He was the one he whispered about when the old ghosts came clawing through his fractured memories. “The one they locked away. The one even I wasn’t allowed to see.”

They started with the basics: a perfected version of the Super Soldier Serum. Not the knockoffs that littered the black market. Not the diluted trash the Flag Smashers used. No, this was the pure, concentrated essence of bioengineered physical supremacy. It made him fast. Strong. Deadly. But that wasn’t enough. HYDRA didn’t want a man—they wanted a god.

They replaced his bones with vibranium, stolen from the very heart of Wakanda in a mission so secret even the Dora Milaje never learned of it. His skeleton was a lightweight fortress, a perfect balance between flexibility and unbreakability. He could be shot point-blank with an anti-tank rifle and not flinch. He could leap from ten thousand feet and land without cracking a toe. His spine alone was stronger than most armored vehicles.

They burned out his organs, one by one, replacing them with biochemical synth-constructs, living machines that pulsed with a power that didn’t belong in the realm of science. His lungs filtered radiation. His kidneys could process raw acid. His stomach could digest metal. Disease didn’t touch him. Poisons turned inert inside him. He didn’t age. He didn’t sleep. He didn’t need to.

His blood… wasn’t blood. It shimmered when it moved. Viscous and luminous, like glowing starlight mixed with oil. Warm, but synthetic. Slick, but alive. It wasn’t just Extremis. It wasn’t just ectoplasm. It was something else entirely. Something that hummed when it moved, that responded to emotion, that sparked with eldritch light when he was angry. It healed him before injury even registered. It whispered to him in languages he never learned but somehow knew. It could ignite with a thought and turn his veins into conduits of fire and ice and terror. They bled him once, just to see what would happen. The blood ate through the floor, hissed like a serpent, and disappeared through the cracks. The lab tech who performed the procedure dissolved within thirty seconds.

And then there was his skin. It was soft, warm, perfectly human. If you touched him, he felt like a boy in his late teens—young, firm, deceptively fragile. But beneath that flawless layer of polymer-fused dermal tissue was something that didn’t burn, didn’t freeze, didn’t shatter. He walked through fire. He dove into the Mariana Trench. He stood unflinching beneath arctic storms and tropical cyclones. He once fought a vibranium-clawed assassin barehanded and didn’t bleed. The assassin didn’t survive.

But the worst part—what made him truly unkillable—was his heart and his brain.

They didn’t understand what they’d done. HYDRA liked to pretend they were gods, but even gods get scared when they tamper with forces they don’t understand. His heart wasn’t just a pump anymore—it was a fusion of quantum mechanics, biomechanical tubing, and something that throbbed with ectoplasmic radiation. It pulsed at its own rhythm, immune to external manipulation. It couldn’t be stopped. You could shoot him in the chest, burn him to ash, decapitate him—and the heart would keep beating. Worse, it could restart him.

The brain was worse. They hadn’t just enhanced his intelligence. They hadn’t just implanted neural tech and a language matrix and memories from assassins, soldiers, pilots, hackers, spies. No. They’d opened a door in his mind. They’d let something in. Something ancient. Something not from this world. Something not even from this dimension. It whispered to him when the moon was full. It guided his hands during missions. It told him where to strike, who to kill, what to become. Sometimes he heard it laughing.

Sometimes he laughed with it.

Wraith was the culmination of every evil science, every secret experiment, every whispered nightmare stitched together into a boy-shaped thing that wore a black suit and a bored expression and had a voice so calm it made seasoned killers nervous. He could walk into a room, look at you with those sky-blue eyes, and make your heart stop—because something about him was wrong. Not obviously wrong. Not monstrous or alien or robotic. No. It was subtle. A slowness to his smile. A tilt to his head. A precision to his movements that screamed in the back of your brain: This isn’t human. This is pretending to be human.

He escaped, of course. Nothing like him could be contained forever. The facility was a ruin within minutes. Bodies left stacked like cordwood. Walls melted. Floors cracked open. Not even the cameras could capture his escape—the footage was corrupted by a static that made your teeth ache and your eyes bleed. Every hard drive in the facility burned itself from the inside out. There was no trace of the boy they once called Danny Fenton.

Now, there are sightings. Rumors. Whispers. In Madripoor, they say he took down a cartel by himself, and the sky turned green when he screamed. In New York, people say he walked past the Sanctum Sanctorum and Doctor Strange flinched like he’d seen death. Wakandan scouts report strange readings near vibranium deposits—heat signatures that vanish into thin air. S.H.I.E.L.D. has classified him as an Omega-level threat.

The Winter Soldier? He saw him once. In an alley in Prague. Wraith didn’t attack. Didn’t speak. Just stared at him with those glacial eyes before disappearing in a flicker of light that bent reality itself. He didn’t sleep for three days after. When asked what was wrong, he just whispered, “They built something worse than me. And it remembers everything.”

Maybe there’s still a boy inside him, buried under steel and fire and ectoplasm and pain. Maybe that boy is screaming. Maybe he’s plotting. Maybe he’s just waiting. After all, you don’t build something like Wraith and expect him to stay still. You don’t break a boy into a god and expect him to forget.

[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.