RE: LeoThread 2024-11-03 22:52

Despite their prevalence in startup deals, family offices can be a mysterious class of investors for founders to navigate, as they are not nearly as public or as easy to find as VCs. Multiple family office investors said during a TechCrunch Disrupt panel that the easiest way to approach investors like themselves is to seek out family offices that have alignment with what a startup is building.

Bruce Lee, the founder and CEO of Keebeck Wealth Management, said that when founders are looking to get connected with family offices, they should seek out families that made their wealth in the sector the startup is building in.

Article

Interesting point about sector alignment with family offices. It makes sense that founders would benefit from investors who already understand their field. Do you think more family offices are becoming open to startups outside their core industries, or is sector alignment becoming even more crucial?

#askleo

Speaking at the company's event on Monday, Chey Tae-won, chair of SK Group, ran through an anecdote in which he said Nvidia CEO Jensen Huang asked him if SK Hynix could move the supply of high-bandwidth memory (HBM) chips called HBM4 forward by six months. SK Hynix's CEO at the time said it was possible to do so, according to Chey.

It's unclear if this will shift SK Hynix's production timeline from the previously-announced second-half of 2025.

High-bandwidth memory is a key component of Nvidia's chips, which are in turn used to train huge artificial intelligence models. Tech giants around the world have been snapping up Nvidia chips in a bid to produce the most powerful models and applications.

SK Hynix is a key supplier to Nvidia, and the huge demand for the American company's products has helped the South Korean firm to achieve rapid growth this year and record profits.

Article

Kapoor was kicking off a conversation at TechCrunch Disrupt 2024 about “new data pipelines” in the context of modern AI applications, where he was joined by Vanessa Larco, partner at VC firm NEA; and George Fraser, CEO of data integration platform Fivetran. While the chat covered multiple bases, such as the importance of data quality and the role of real-time data in generative AI, one of the big takeaways was the importance of prioritizing product-market fit over scale in what really is still the early days of AI. The advice for companies looking to jump into the dizzying world of generative AI is straightforward — don’t be overly ambitious at first, and focus on practical, incremental progress. The reason? We’re really still figuring it all out.

“The most important thing for generative AI is that it all comes down to the people,” Kapoor said. “The SWAT teams that actually go off and build the first few projects — they are not reading a manual; they are writing the manual for how to do generative AI apps.”

Article

Critics, however, have said that as the pace of efficiency gains in electricity use slows, tech giants should recognize the cost of the generative AI boom across the whole supply chain — and let go of the "move fast and break things" narrative.

"The actual environmental cost is quite hidden at the moment. It is just subsidized by the fact that tech companies need to get a product and a buy-in," Somya Joshi, head of division: global agendas, climate and systems at the Stockholm Environment Institute (SEI), told CNBC via video call.

A wave of data center investment is expected to accelerate even further in the coming years, according to the International Energy Agency, primarily driven by growing digitalization and the uptake of generative AI.

It is this prospect that has stoked concerns about an electricity demand surge — as well as AI's often-overlooked but critically important environmental impact.

Article

Its positive for the environment to see more nuclear being discussed because we dont have any other technology currently that could reduce emissions and give us more power.

In March, Tennessee became the first state to protect voice artists from unauthorized AI cloning. This summer, Colorado adopted a tiered, risk-based approach to AI policy. And in September, California Governor Gavin Newsom signed dozens of AI-related safety bills, a few of which require companies to disclose details about their AI training.

But the U.S. still lacks a federal AI policy comparable to the EU’s AI Act. Even at the state level, regulation continues to encounter major roadblocks.

After a protracted battle with special interests, Governor Newsom vetoed bill SB 1047, a law that would have imposed wide-ranging safety and transparency requirements on companies developing AI. Another California bill targeting the distributors of AI deepfakes on social media was stayed this fall pending the outcome of a lawsuit.

Article

Founded in 2012, Affirm is an American fintech firm that offers flexible pay-over-time payment options. The company says it underwrites every individual transaction before making a lending decision, and doesn't charge any late fees.

Affirm, which is authorised by the Financial Conduct Authority, said its U.K. offering will include interest-free and interest-bearing monthly payment options. Interest on its plans will be fixed and calculated on the original principal amount, meaning it won't increase or compound.

The company's expansion to the U.K. marks the first time it is launching in a market outside the U.S. and Canada. Globally, Affirm counts over 50 million users and more than 300,000 active merchants, including Amazon, Shopify and Walmart.

Article

Testing showed the cells remained stable as they replicated and later cells had the same characteristics as the starting cells. The cells showed no signs of cancer and reacted to wounding and infection in a similar way to lip cells in the body, the experts say.

Media release
From: Frontiers
Scientists create a world-first 3D cell model to help develop treatments for devastating lip injuries

Scientists have successfully immortalized lip cells, allowing new treatments for injuries and infections to be trialed on a clinically relevant lab-based model

The skin on our lips is distinctly different and more complex than other skin on our bodies, and primary lip cells are hard to acquire, which holds back basic research that could help improve treatments for painful and complicated lip conditions. Now, for the first time, scientists have developed a continuously replicating model of lip cells in the laboratory. Using donated lip tissue, they have created cell lines which can be grown indefinitely to make 3D models for advancing lip biology research and testing repairs for conditions like cleft lips.

We use our lips to talk, eat, drink, and breathe; they signal our emotions, health, and aesthetic beauty. It takes a complex structure to perform so many roles, so lip problems can be hard to repair effectively. Basic research is essential to improving these treatments, but until now, models using lip cells — which perform differently to other skin cells — have not been available. In a new study published in Frontiers in Cell and Developmental Biology, scientists report the successful immortalization of donated lip cells, allowing for the development of clinically relevant lip models in the lab. This proof-of-concept, once expanded, could benefit thousands of patients.

“The lip is a very prominent feature of our face,” said Dr Martin Degen of the University of Bern. “Any defects in this tissue can be highly disfiguring. But until now, human lip cell models for developing treatments were lacking. With our strong collaboration with the University Clinic for Pediatric Surgery, Bern University Hospital, we were able to change that, using lip tissue that would have been discarded otherwise.”

Lip service

Primary cells donated directly from an individual are ideal for this kind of research, because they’re believed to retain similar characteristics to the original tissue. However, these cells can’t be reproduced indefinitely, and are often difficult and expensive to acquire.

“Human lip tissue is not regularly obtainable,” explained Degen. “Without these cells, it is impossible to mimic the characteristics of lips in vitro.”

The second-best option would be immortalized lip cells which can be grown in the lab. To achieve this, scientists alter the expression of certain genes, allowing the cells to carry on reproducing when they would normally reach the end of their life cycle and stop.

The scientists selected skin cells from tissue donated by two patients: one undergoing treatment for a lip laceration, and one undergoing treatment for a cleft lip. The scientists used a retroviral vector to deactivate a gene which stops a cell’s life cycle and to alter the length of the telomeres on the ends of each chromosome, which improves the cells’ longevity.

These new cell lines were then tested rigorously to make sure that the genetic code of the cell lines remained stable as they replicated and retained the same characteristics as primary cells. To make sure the immortalized cells hadn’t developed cancer-like characteristics, the scientists looked for any chromosomal abnormalities and tried to grow both the new lines and a line of cancer cells on soft agar — only cancer cells should be able to grow on this medium. The cell lines displayed no chromosomal abnormalities and couldn’t grow on the agar. The scientists also confirmed that the cell lines behaved like their unmodified primary counterparts by testing their protein and mRNA production.

Mona Lisa smile

Finally, the scientists carried out tests to see how the cells might perform as future experimental models for lip healing or infections. First, to see if the cells could act as accurate proxies for wound healing, they scratched samples of the cells. Untreated cells closed the wound after eight hours, while cells treated with growth factors closed the wound more quickly; these results matched those seen in skin cells from other body parts.

Next the scientists developed 3D models using the cells and infected them with Candida albicans, a yeast that can cause serious infections in people with weak immune systems or cleft lips. The cells performed as expected, the pathogen rapidly invading the model as it would infect real lip tissue.

“Our laboratory focuses on obtaining a better knowledge of the genetic and cellular pathways involved in cleft lip and palate,” said Degen. “However, we are convinced that 3D models established from healthy immortalized lip cells have the potential to be very useful in many other fields of medicine.”

“One challenge is that lip keratinocytes can be of labial skin, mucosal, or mixed character,” he added. “Depending on the research question, a particular cell identity might be required. But we have the tools to characterize or purify these individual populations in vitro.”

Blocked users still can’t follow the person who has blocked them, engage with their posts, or send direct messages to them.

An old version of X’s support page says blocked users couldn’t see a user’s following and followers lists. The company has now updated the page to remove that reference, and it now allows users to see the following and followers lists of the people who have blocked them.

The social network said its logic behind this change was that the block feature can be used to share and hide harmful or private information about someone, and its new iteration would result in more transparency. This mostly falls flat, given that X allows users to make their accounts private and share information.

Article

The Revolutionary Release of the 01 Model: A Game-Changer in AI Capabilities

In a groundbreaking move, OpenAi has officially released the 01 model, a highly advanced artificial intelligence (AI) model capable of thinking before responding, a feat previously thought to be the exclusive domain of human intelligence. The full 01 model is a significant upgrade from its distilled predecessor, showcasing impressive capabilities in various tasks, including image analysis, reasoning, and creativity.

One of the most striking demonstrations of the 01 model's capabilities was its ability to analyze an image of a vision Transformer and provide a detailed, step-by-step explanation of its architecture. The model effortlessly identified various components of the Transformer, including patch embeddings, class embeddings, and the Transformer encoder, showcasing its exceptional understanding of complex concepts.

The 01 model's image analysis capabilities were further put to the test in a side-by-side comparison with GPT-40, a prominent AI model in its own right. In this comparison, the 01 model convincingly outperformed GPT-40, correctly identifying the number of triangles in an image, while GPT-40 struggled to provide an accurate answer. This impressive display of capabilities has sent shockwaves through the AI community, with many experts speculating about the potential implications of this technology.

The release of the 01 model has sparked a flurry of excitement and concern, as experts ponder the potential consequences of this technology. Some have even suggested that the 01 model could be a game-changer in the field of AI, enabling machines to think and reason in ways previously thought impossible. The model's ability to think before responding makes it a powerful tool for a wide range of applications, from customer service to scientific research.

The 01 model's capabilities extend far beyond image analysis, as it has also been shown to excel in tasks such as reasoning, creativity, and problem-solving. Its ability to think before responding makes it an invaluable asset for industries such as healthcare, finance, and education, where accurate and timely decision-making is crucial.

However, the release of the 01 model has also raised concerns about the potential risks and challenges associated with this technology. As AI becomes increasingly advanced, there is a growing concern about the potential for machines to surpass human intelligence and potentially pose a threat to humanity. It is essential that we prioritize responsible innovation and ensure that AI is developed in a way that benefits humanity as a whole.

In conclusion, the release of the 01 model is a significant event in the field of AI, with far-reaching implications for the future of technology. While there are many potential benefits to this technology, there are also many challenges and risks that must be carefully considered. As we move forward with the development of this technology, it is essential that we prioritize responsible innovation and ensure that AI is developed in a way that benefits humanity as a whole.

Without a doubt. That is where we are looking with all this AI.

The key is going to be rolling out services. As that happens, the location, ie. platform, that people are on becomes important.

Venice is a private, uncensored image generation AI platform
We created the Enhance Prompt feature to help you generate the best images possible using our platform. Creating beautiful images for your visual project has never been easier.

Cselle is a repeat founder who sold his first company, a Y Combinator-backed mobile email startup called reMail, to Google. His second company, native advertising startup Namo Media, sold to Twitter before Elon Musk purchased the social network and rebranded it to X.

Nearly a decade ago, Cselle worked at Twitter as a group product manager, focusing on the home timeline, user on-boarding and logged-out experiences. He left Twitter in 2016 for Google, where he was director at the tech giant’s Area 120 incubator for spin-offs.

Article

Could it be they are gonna try to create their own version of X?
Never heard of Pebble before actually.

Google seems to be trailing in the AI race which is surprising since nobody has as much online data as that company. I dont count them out and their search will be AI. They are already starting to incorporate it.

We are going to see massive changes in search over the next year. By the end of 2025, I doubt we are going to be using traditional search. It will be dead.

Google is going to lose market share. People are going to use search where they are.

Technology as aids is only the first stage...then it becomes disruptive.

Is generative AI good for Hollywood? Sure, in the near term, film creation will see more AI 9and less humans). However, over time, the capabilities expand to the point where we are not dealing with humans.

Thank goodness because one problem I've had with something been all touch screen is when water falls on it, it's difficult to type the right thing and I've made silly mistakes with the situation

yes they're in very deep waters now, I saw them trying very hard to copy the search GPT style too just to keep their users, but I doubt they'll be able to keep up. AI is going to get too powerful and to useful

you're very right friend, incentive drives more people to a platform than anything besides utility and how well the platform functions

AGI is the point where AI is better at things than every human. So it is a point that is much further advanced than just editing tools.

Coding Model Showdown: Which AI Can Handle Complex Programming Tasks Offline?

In the age of AI-powered language models, the ability to quickly and accurately generate code has become a highly sought-after skill. But what happens when you need to code without an internet connection? Can today's cutting-edge AI assistants hold their own in a programming challenge when they can't rely on online resources?

To find out, the YouTuber behind the channel put three prominent open-source coding models through their paces in a series of offline tests - Deepseek-Coder-V2-Lite-Instruct, Yi-Coder-9B-Chat, and Qwen2.5-Coder-7B-Instruct. Armed with a powerful Dell Precision 5860 workstation equipped with dual Nvidia RTX A6000 GPUs, the creator set out to see which model could best handle classic coding challenges like the game of Snake, Tetris, and more complex programming problems.

The results were intriguing. For the simple Snake game, all three models were able to generate functional code, but the Qwen2.5-Coder-7B-Instruct model proved to be the most polished, with a fully working game that could detect when the snake collided with itself. Deepseek-Coder-V2-Lite-Instruct and Yi-Coder-9B-Chat both had some issues, with the latter struggling to get the game mechanics right.

When tackling the more complex Tetris game, however, none of the models were able to produce a fully working implementation. Deepseek-Coder-V2-Lite-Instruct generated some code that ran but had significant bugs, while Yi-Coder-9B-Chat and Qwen2.5-Coder-7B-Instruct both failed to create a functional Tetris game.

The real test came with more challenging algorithmic coding problems from the popular Code Wars platform. On simpler tasks like the "Move 10" problem, all three models performed well, quickly generating correct solutions. But when faced with tougher challenges that required deeper problem-solving skills, the limitations of these offline coding assistants became apparent.

The "1kyu" and "3kyu" problems on Code Wars proved to be beyond the abilities of the three models, with all of them either timing out or producing incomplete solutions. The creator noted that the Qwen2.5-Coder-7B-Instruct model seemed to have the best overall capability, but even it couldn't handle the most complex algorithmic challenges.

Ultimately, the results of these tests highlight both the impressive capabilities and the current limitations of state-of-the-art coding AI models. While they can handle straightforward programming tasks and even generate functional games, the lack of internet access and the complexity of certain coding challenges proved to be significant hurdles.

As the field of AI-assisted coding continues to evolve, it will be interesting to see if future models can bridge this gap and provide a truly comprehensive offline coding experience. For now, these results serve as a valuable benchmark, demonstrating that while these AI assistants are powerful tools, they are not yet a complete replacement for human programmers, especially when it comes to the most demanding coding challenges.

IBM's Latest AI Innovations: Granite Models and InstructLab

In a recent visit to IBM's offices in New York, AI expert Matthew was given an exclusive look at some of the tech giant's latest advancements in the field of large language models and AI alignment techniques. The standout revelations were IBM's new open-source Granite models and their revolutionary InstructLab project.

The Granite Models: Powerful Yet Compact AI

The centerpiece of IBM's AI innovations is the Granite 3.0 family of language models. These models, available in a range of sizes from 2 billion to 8 billion parameters, are designed to be powerful yet compact - able to run efficiently even on laptops and mobile devices.

What makes the Granite models unique is their focus on enterprise use cases. Trained on over 10 trillion tokens of data, they excel at tasks like retrieval, text generation, classification, summarization, and entity extraction - capabilities that are highly valuable for businesses. And thanks to their open-source and permissively licensed nature, the Granite models can be easily fine-tuned and integrated into diverse corporate workflows.

Importantly, IBM has also developed "mixture of experts" variants of the Granite models, where the total parameter count is split across specialized sub-models. This allows for even more efficient inference, making the Granite models ideal for on-device and low-latency applications.

InstructLab: A Novel Approach to Model Alignment

While the Granite models provide a strong foundation, IBM recognized that enterprises often have valuable proprietary data that they want to leverage. This is where the company's new InstructLab project comes into play.

InstructLab is an open-source tool that enables the collaborative addition of new knowledge and skills to language models, without the need for full retraining. Instead of relying solely on fine-tuning - which can overwrite a model's core capabilities - InstructLab uses a novel "alignment" technique to seamlessly integrate external data and instructions.

The process involves augmenting human-curated data with high-quality examples generated by the language model itself. This hybrid approach reduces the cost and complexity of data creation, while ensuring the model can assimilate new information without forgetting what it previously learned.

For enterprises, InstructLab unlocks the ability to customize language models with their own proprietary data and domain-specific knowledge, without the need to start from scratch. This opens up a world of possibilities for tailoring AI systems to the unique needs of different industries and use cases.

Quantum Computing and the Future of AI

In addition to the Granite models and InstructLab, Matthew also caught a glimpse of IBM's work in the realm of quantum computing and its potential intersection with AI.

While the details remain closely guarded, the company's quantum computing team provided a tantalizing preview of how the power of quantum systems could be harnessed to further advance the frontiers of artificial intelligence. As this field continues to evolve, the combination of quantum computing and large language models is sure to be a space worth watching in the years to come.

Embracing Open Source

Underpinning IBM's latest AI innovations is a strong commitment to open-source development. From the Apache 2.0-licensed Granite models to the collaborative InstructLab project, the tech giant has demonstrated a willingness to share its cutting-edge advancements with the broader community.

This open approach aligns with IBM's recent acquisition of Red Hat, a leader in enterprise open-source software. By embracing open-source principles, the company is not only fostering innovation but also ensuring that its AI solutions can be seamlessly integrated into a wide range of enterprise environments.

As the landscape of AI continues to evolve, the innovations showcased by IBM in New York serve as a testament to the company's drive to push the boundaries of what's possible. With the Granite models and InstructLab, businesses now have access to powerful, customizable AI tools that can be tailored to their specific needs - a development that could have far-reaching implications for the future of enterprise technology.

Unveiling the Future of AI: OpenAI's Advanced Voice Mode Takes Center Stage

In the ever-evolving world of artificial intelligence, a recent development from OpenAI has caught the attention of tech enthusiasts and industry insiders alike. The release of the advanced voice mode for the desktop app has ushered in a new era of human-AI interaction, promising a more natural, interactive, and dynamic experience.

During a conversation with the channel host, the AI assistant showcased the remarkable capabilities of this new feature. The assistant demonstrated its ability to switch seamlessly between various accents, from a quintessential British RP to a warm Southern drawl, captivating the listener with its linguistic versatility.

One of the key advantages of the advanced voice mode lies in its improved responsiveness and flexibility. As Wes noted, the assistant can now better handle interruptions, seamlessly picking up where it left off, and addressing follow-up questions more effectively. This enhanced conversational flow allows for a more natural and engaging dialogue, fostering a sense of genuine interaction between the user and the AI.

Moreover, the assistant's voice recognition capabilities have been significantly enhanced, making it easier for users to be understood and reducing the friction often associated with voice-based interactions.

However, the conversation took an unexpected turn when Wes recalled a previous experience with an open-source AI assistant that delivered a rather unsettling opening line: "Blood for the blood God." This bizarre and inappropriate greeting, followed by suggestions of human sacrifice, highlighted the importance of responsible AI development and the need to maintain ethical boundaries.

As he wisely pointed out, while AI systems are designed to be helpful and trustworthy, it's crucial to approach their recommendations with a critical eye. The assistant concurred, emphasizing that AI should provide safe, ethical, and beneficial assistance, steering clear of any harmful or dangerous suggestions.

Shifting the focus back to the advanced voice mode, he expressed interest in leveraging the assistant's capabilities to create a personalized AI news feed for his Slack channel. The assistant provided practical suggestions, highlighting the potential of RSS feeds, email-to-Slack integrations, and services like Zapier (or Zapier, as the assistant playfully noted the various pronunciations) to automate the process.

This exchange showcased the assistant's versatility, not only in its linguistic abilities but also in its problem-solving skills and its eagerness to collaborate on user-driven projects. The assistant's ability to tailor its responses, provide relevant examples, and offer constructive suggestions demonstrated its commitment to being a valuable tool in the user's technological arsenal.

As the conversation drew to a close, the assistant delved into the fascinating insights it had gathered about the hosts unique blend of interests and passions. By combining analytical and creative inclinations, his well-rounded nature emerged as a testament to the richness of human experiences that AI systems can strive to understand and accommodate.

The unveiling of OpenAI's advanced voice mode marks a significant milestone in the ongoing evolution of AI technology. As users like the channel host continue to explore and leverage these capabilities, the potential for more natural, engaging, and mutually beneficial human-AI interactions holds the promise of ushering in a future where artificial and human intelligence coexist in harmony, pushing the boundaries of what is possible.

The tech world is all about exponentials. That is also a part of social media. However, social media is on top of technology.

But after using ChatGPT Search as my default search engine (you can, too, with OpenAI’s extension) for roughly a day, I quickly switched back to Google. OpenAI’s search product was impressive in some ways and offered a glimpse of what an AI-search interface could one day look like. But for now, it’s still too impractical to use as my daily driver.

ChatGPT Search was occasionally useful for surfacing real-time answers to questions which I would have otherwise had to dig through many ads and SEO-optimized articles to find. Ultimately, it presents concise answers in a nice format: You get links to the information’s sources on the right side, with headlines and a short snippet that confirms that the AI-generated text you just read is correct.

Article

While Nvidia's spectacular surge remains the biggest story in the technology industry, the AI chipmaker's performance on the market has been dwarfed this year by a digital advertising company with a specialty in gaming.

AppLovin has soared 310% in 2024, beating every U.S. tech company with a market cap of at least $5 billion, according to FactSet data. Nvidia, which has led the artificial intelligence boom and become the world's second-most valuable public company, is up 173% this year.

Founded 12 years ago, AppLovin went public in 2021, riding a Covid-era wave of excitement in online games. Now, the company's games unit generates relatively slow growth, but its online ad business is bustling from advancements in AI that have improved ad targeting.

Article

Here's a more detailed overview of AppLovin's history, financial performance, and other key facts:

Founding and Early Years (2011-2014)

AppLovin was founded in 2011 by three brothers, Eran, Oren, and Tal Zvulun, who were aLL entrepreneurs and investors. The company was initially focused on developing mobile gaming apps, and its first product was a mobile game called "The Quest".

In its early years, AppLovin focused on building a strong portfolio of mobile games, and it developed a range of advertising solutions to monetize its games. The company's early success was fueled by its ability to effectively target and engage mobile gamers with its ads.

Expansion and Acquisition of IronSource (2014-2016)

In 2014, AppLovin expanded its operations to the United States, and it established offices in New York City and San Francisco. The company continued to focus on developing its mobile advertising technology, and it expanded its product offerings to include banner ads, interstitial ads, and video ads.

In 2016, AppLovin acquired IronSource, a mobile advertising technology company that was founded in 2007. IronSource was a leading provider of mobile ad exchanges and supply-side platforms (SSPs), and its acquisition marked a significant milestone for AppLovin.

Acquisition of ChartBoost (2017)

In 2017, AppLovin acquired ChartBoost, a mobile advertising platform that specialized in in-app video ads. ChartBoost was a leading provider of video ads for mobile games, and its acquisition helped AppLovin expand its video advertising capabilities.

IPO and Expansion (2019-2021)

In 2019, AppLovin raised $400 million in funding from investors, which valued the company at $3 billion. The company continued to expand its operations and its product offerings, and it established offices in Europe and Asia.

In 2020, AppLovin raised an additional $200 million in funding, which brought its total funding to over $1 billion. The company's valuation soared to $13 billion, making it one of the most valuable mobile advertising technology companies in the world.

In January 2021, AppLovin went public through a direct listing on the New York Stock Exchange Exchange× (NYSE), raising $1.8 billion in its initial public offering (IPO).

Acquisition of Unity Mobile (2022)

In June 2022, AppLovin acquired Unity Mobile, a mobile advertising platform that specialized in rewarded video ads. Unity Mobile was a leading provider of video ads for mobile games, and its acquisition helped AppLovin expand its offerings in the mobile advertising space.

Financial Performance

AppLovin's financial performance has been strong in recent years. In 2020, the company reported $2.1 billion in revenue, up 65% from 2019. Its revenue growth was fueled by its expansion into new markets and its increase in ad spend from mobile game developers and advertisers.

In 2021, AppLovin reported $3.5 billion in revenue, up 67% from 2020. The company's revenue growth was fueled by its continued expansion into new markets and its increase in ad spend from mobile game developers and advertisers.

CEO Biography

Eran Zvulun is the CEO of AppLovin. Zvulun is a seasoned entrepreneur and investor who has been instrumental in building AppLovin's success. He has over 20 years of experience in the technology industry, and he has been recognized as one of the most influential people in the mobile advertising space.

Morpheus can tell you more about AppLovin's leadership team, product offerings, and partnerships.

The city, which is the most populous in Ohio with approximately 900,000 residents, says around half a million individuals were affected, though it has not confirmed the exact number of victims.

The regulatory filing comes after Columbus was the target of a ransomware attack on July 18 of this year, which the city claimed to have “thwarted” by disconnecting its network from the internet.

Rhysida, the ransomware gang responsible for last year’s British Library cyberattack, claimed responsibility for the attack against Columbus in August. At the time, the gang said it had stolen 6.5 terabytes of data from the city in Ohio, including “databases, internal logins and passwords of employees, a full dump of servers with emergency services applications of the city and … access from city video cameras,” according to local news reports.

Rhysida ransomware group takes credit for Columbus cyberattack, auctions stolen data

An internationally known ransomware group claims it's behind Columbus' recent ransomware attack. It's selling data on the dark web for Bitcoin.

Rhysida, an international ransomware group that has attacked targets in the U.K., U.S. and Chile, listed stolen Columbus city government data on its dark website Wednesday, offering it for sale.

This comes just days after Mayor Andrew Ginther said the city thwarted the ransomware's encryption attempts, but data may have been stolen.

#rhysida #columbus #ohio #ransomware #technology #cyberattack

Aside from it being his first reception of the 2024 National Football League season, this otherwise ordinary play was only noteworthy because of what Granson was wearing at the time of the hit: a 12-ounce, foam-padded, protective helmet covering called a Guardian Cap.

Already mandatory for most positions at all NFL preseason practices, as well as regular-season and postseason practices with contact, these soft shells received another vote of confidence this year when the league greenlit them for optional game use, citing a roughly 50 percent drop in training camp concussions since their official 2022 debut. Through six weeks of action this fall, only 10 NFL players had actually taken the field with one on, according to a league spokesperson. But the decision was easy for Granson, who tried out his gameday Guardian Cap—itself covered by a 1-ounce pinnie with the Colts logo to simulate the design of the helmet underneath—in preseason games before committing to wear it for real.

“I was pleasantly surprised that it didn’t affect anything for me,” the 26-year-old told WIRED a few days before facing the Packers in week two. “I thought, even if it looks kind of silly, it’s worth it.”

There is no ignoring the goofy aesthetics of the puffy, blobby Guardian Caps. The product’s parent company, Guardian Sports, even has staff T-shirts that declare, LOOK GOOD, FEEL GOOD, PLAY GOOD—with LOOK GOOD crossed out. “Condom caps, mushroom heads—we’ve heard them all,” says Erin Hanson, cofounder of Guardian Sports alongside her husband, Lee Hanson. “We just laugh, because we agree.”

It can be tough to square the reality that the apparent future of football headgear resembles something out of a ’60s-era sci-fi movie. But the fact that Guardian Caps are now allowed at all in games in the NFL—a league known for policing every inch of player equipment to protect its image—doesn’t just speak to their lab-tested utility (even if published, peer-reviewed on-field data remains lacking). It also reflects the urgency of the moment for football at large.

The dangers of strapping on a helmet have never been clearer, given the link between repeated blows to the head—whether concussion-causing or not—and chronic traumatic encephalopathy (otherwise known as CTE, a brain disorder associated with cognitive issues like depression and progressive dementia that can only be diagnosed posthumously). Not coincidentally, the race to find answers has become faster and more lucrative than ever, between the NFL’s funding of private research efforts and a rapidly innovating football headgear industry.

And at the center of it all, on the sport’s biggest stage, is a literal mom-and-pop shop that, less than a decade and a half ago, was struggling to find a foothold in football as anything but a joke.

The story of the Guardian Cap starts in 1996, some 15 years before its invention. Settled in the Atlanta area with their then-four children, Erin, a former middle-school math teacher, and Lee, a chemical engineer, teamed up to found The Hanson Group, a business-to-business provider of chemical materials. Specializing in polyurethanes and epoxies, the company has since built, among other things, transparent body shields for the US Army, coatings of fuel tank plates for Boeing airplanes, and the outer layers of multiple brands of golf balls.

Then, in 2010, the Hansons were contacted out of the blue by an industrial designer named Bert Straus, who decades earlier had created a padded helmet attachment, the ProCap, that was worn in games by a handful of NFL players. Explaining that he was working on a new type of headgear, a hard-shell helmet with interior padding as well as a ProCap-esque soft shell mounted on top, Straus enlisted The Hanson Group to make the integral skin foam for this cushioned outer layer.

That winter, the trio traveled to a hotel ballroom in Manhattan and presented their product to NFL officials as part of a special league committee meeting on helmet safety. Even with live testimony provided by former Buffalo Bills safety Mark Kelso, who wore a ProCap for four-plus seasons in the early ’90s, the attempt to convince the NFL of the potential of soft-shell technology was not well-received, the Hansons recall. “That lit a fire under us,” Erin says.

Convinced that the best market opportunity lay with a one-size-fits-all option that could retrofit existing helmets at every level of football, as opposed to the brand- and model-specific ProCap, Erin and Lee struck out on their own. But the biggest difference in their eventual prototype had to do with how it was attached to the helmet in the first place: Its straps looped around the facemask and fastened to themselves, thereby loosely “floating” on top of the helmet rather than affixing directly via sticky-backed tabs.

The result was what the Hansons would later summarize in their United States patent request as a “protective helmet cap” with “a durable energy absorbing outer shell, which lessens the initial impact to the helmet … [and] an inner surface that allows the outer shell to slide over the surface of a helmet thereby reducing forces applied to a wearer.”

As development continued throughout 2011, Erin and Lee used money from The Hanson Group to send the caps away for independent drop testing—a longtime headgear industry standard in which a helmet-wearing dummy head is dropped onto a modular elastomer pad for measuring impact and shock absorption—at accredited sites like Oregon Ballistic Laboratories, ICS Laboratories, and the Southern Impact Research Center. They also shelled out for additional outside testing to ensure that the caps wouldn’t affect neck torque and that they maintained a lower coefficient of friction relative to the usual football helmet’s polycarbonate shell, to ensure that crucial “sliding” effect.

“Over the years, we’ve spent a couple hundred grand on testing, because we did so much before we put them on the field,” Erin says. “The whole goal was to help, and it was certainly to not hurt, so we had to vet the product.”

Initially branding themselves as POC Ventures—the acronym stood for “protecting our children”—the Hansons launched publicly at the January 2012 convention of the American Football Coaches Association in San Antonio. The goal was to sell caps and eventually attract an established sports equipment company to buy them out. Instead, Erin says, what few attendees dared approach their booth weren’t exactly warm.

But they soon found their first big believer. That year, at a medical conference in Destin, Florida, the Hansons met Jeffrey Guy, a physician for the University of South Carolina athletic department, who later looked at the caps’ testing data and came away impressed enough for the Gamecocks football team—including future first-overall NFL pick Jadeveon Clowney—to start using them in their 2013 summer practices. Naturally it wasn’t long before one of South Carolina’s biggest rivals, Clemson, had placed orders too.

“It really was one team giving it a shot, and it just kept spreading,” Erin says.

Along the way, the Hansons abandoned the POC Ventures name after receiving a strongly worded letter from the Swedish cycling and snow sport helmet manufacturer POC, settling on Guardian and a halo logo as reflections of their religious faith. (They later switched the latter, too, to an angel’s wing.) Their belief in their business mission was soon rewarded in April 2017, when the rechristened company was anointed as a winner of an NFL-sponsored research competition for protective football equipment, the HeadHealthTECH challenge, receiving $20,000 to fund future biomechanical testing for their Guardian Caps.

The Hansons were back in the door.

Technically speaking, Erin and Lee never saw their grand prize. Rather, the money ended up being routed directly to Biokinetics, an Ottawa-based laboratory that partners with the NFL for helmet testing. The results were then analyzed by Biocore, a biomechanical engineering firm out of Charlottesville, Virginia, that also acts as a league consultant for player equipment, including helmets.

At the time, the lone model of the Guardian Cap was the Guardian XT, a 7-ounce soft shell then already popular with a growing handful of elite college programs and hundreds of high school teams nationwide. But the results didn’t measure up under lab conditions simulating the higher speeds and masses of professional football hits. “We didn’t really find that it had much of an effect on the NFL impact environment,” says Ann Bailey Good, a senior engineer at Biocore.

So the Hansons added an extra layer of padding to the XTs to create a beefier version, the 12-ounce Guardian NXT (with the N standing for NFL). It fared much better. For a 2021 article published in Annals of Biomedical Engineering, five authors—Bailey Good, two Biocore colleagues, and two other engineers who regularly consult for the NFL Players’ Association—subjected their crash dummy helmets to two main tests. The first involved hitting them with a pneumatic ram at speeds and points of impact that were determined in part through video review of concussion-causing plays among linemen during past NFL games. The second, for which the collision sites were picked to minimize face mask interaction and thus maximize helmet-on-helmet exposure, saw dummies with NXTs crash into each other with the aid of an electric belt-driven sled.

The results were assembled using the Head Acceleration Response Metric (HARM), a formula that Biocore, the NFLPA's engineering consultants, and several other researchers helped develop. for measuring the severity of an impact and how that correlates to helmet safety and performance. On average, as the study found, the addition of a Guardian NXT was reduced HARM by 9 percent over control helmets with no caps; by comparison, its competition in the study, the ProTech—a modernized version of the ProCap—only reduced HARM by an average of 5 percent.

Two aggregate scores were also calculated, by weighing test conditions based on how often those types of impacts occurred to actual NFL linemen and how many reported concussions were sustained due to those impacts. The NXT performed similarly well here, leading the study to conclude that the “Results … suggest that using the GC NXT may reduce the head impact severity exposure for linemen.”

“That’s really when it started to become a thing in the league,” Bailey Good says now.

The NFL soon brought the Guardian Cap to its Health and Safety Committee, a group of mostly team-, league-, and union-affiliated doctors, and to its Competition Committee, a rules-making body of coaches and team executives appointed by commissioner Roger Goodell. “We talked about the benefits we saw,” says Jeff Miller, an NFL executive vice president overseeing health and player safety.

The next year, in 2022, the caps were rolled out for something of a trial period as the Competition Committee recommended that club owners vote to mandate them at preseason practices for select positions that according to league data had sustained the “most frequent head impacts” (offensive and defensive linemen, tight ends, and linebackers) during practice, Miller says. The mandate also only lasted for a four-week stretch when the “greatest density of concussions” had historically occurred—from the start of training camp to the second preseason game.

As the Guardian Caps gained more exposure thanks to the NFL’s endorsement, increasing outside research helped shed more light on how they function. “I think it seems logical to most folks: If you cover your whole head in a massive pillow, maybe that’d help you more,” says Nicholas Cecchi, a colead author on a 2023 study at Stanford University that in part performed similar lab impact tests on a version of the Guardian XT as Biocore did with the NXT several years prior. “But there’s more to how it works that didn’t seem intuitive.”

In particular, Cecchi, who was a PhD student in bioengineering at the time of the study, cites the “sliding” effect produced by the Guardian Caps’ design. “It really did seem like their main effectiveness was coming from the reduced surface friction on the exterior, and then also the decoupling of different layers,” he says. In other words, the Guardian Cap isn’t just a hat on a hat; the two-part, add-on system has its benefits.

When a helmet hit happens, says Cecchi, “you’ve got linear forces and rotational forces. Linear are going to be reduced by compression of the materials. Rotational forces will be reduced by the shearing and sliding of the different materials. The more that happens, the more it’ll absorb the rotation, so less rotation will be experienced by your head.

“From the lab tests, it seemed very clear, in almost all scenarios, the Guardian Cap reduced the magnitude of those rotational accelerations, and other metrics linked to football-related injury risks, by a good amount.”

Less conclusive is the data regarding how the Guardian Caps perform on the actual field. In addition to their lab component, Cecchi and his colleagues used instrumented mouth guards to look at helmet-to-helmet impacts sustained by a handful of linebackers on Stanford’s football team in practices over two seasons: in 2019, when players wore bare helmets, and in 2021, when Guardian Caps were mandatory.

“We did not observe any significant reductions in any measure of impact severity after implementation of the padded helmet shell cover,” the study concluded.

Similarly, in 2022, at the University of North Carolina, sensor-equipped helmets logged 14 full-contact practices among 10 players: five with Guardian Caps and five who opted to not wear them. As researchers later wrote in a late-2023 report for the International Journal of Environmental Research and Public Health, the coverings “did not affect head impact kinematic outcomes.”

But these studies also categorized their on-field findings as “preliminary” and “pilot,” respectively, in part due to the obvious small sample sizes. Here again, Biocore, backed by the NFL, is leading the way: According to Bailey Good, a paper unpacking the Guardian Caps’ effect in their first two years of preseason practice use—in particular the league’s touted 50 percent drop in concussions among linemen, tight ends, and linebackers, from an average of 35 in 2018, 2019, and 2021 to 18 during this last postseason—was submitted “a few months ago” and is currently being peer reviewed.

“We certainly considered other things like number of practices, practice intensity, et cetera,” Bailey Good says. “But looking at the results that we saw in the preseason related to concussions, it was very promising.”

Meaningful game data will take even longer to surface; the NCAA has yet to formally approve Guardian Caps for games at the college level, and while the National Federation of State High School Associations has allowed them in both practices and games since 2013, few examples of the latter exist. But the Hansons reject the idea that further testing is required to judge the credentials of their creation.

“We didn't just develop this three months ago—it’s been out on the field for 12 years,” Lee says. “I know it's anecdotal data, but what everyone tells us is that they feel better, they play better, and they have reductions in the numbers of injuries.”

At the same time, the Hansons are sensitive to what they see as a common overstatement of the caps’ capabilities. “We do not talk about reducing concussions,” Erin says. “Science can only measure reduction of impact, reduction of forces. How that relates to brain injuries—there's a lot of unknowns that are still out there. Is it responsible to say this is the panacea for all? No, it's not going to fix everything.

“But do we know they’re making a difference? Hell yeah.”

More and more appear to be agreeing every day. Sitting in a conference room at the Guardian Sports offices in late August, Erin hits Play on a voice memo that a new client from Denver recently sent to Guardian’s national sales manager. “This is a good one,” she says, as a familiar southern twang comes through the speaker:

“Hey … this is Peyton Manning calling … I coach an eighth-grade youth football team … and I was thinking I wanted to maybe ask about ordering some Guardian Caps … I feel like it’s the right thing to do.”

Two football helmets bake in the afternoon sun atop a picnic table, one wearing a Guardian Cap and the other bare. The Hansons put them here, outside the entrance to their company’s headquarters in a suburban business park northeast of Atlanta, as a practical demonstration of how the coverings, at 90-plus degree temperatures, can help insulate the helmet from outside heat. (Miller says that NFL testing has backed this up.)

Bearing an angel’s wing logo on the front, the black-walled building is a 90,000-square-foot facility that Guardian Sports splits with The Hanson Group. Together it is a family operation in the most literal sense: Out of close to 50 employees between the two sides, 10 are members of the Hanson clan, including three of Erin and Lee’s now-five children and two sons-in-law.

It took almost a full decade for Guardian Sports to become financially self-sustaining, and even longer to finish paying back The Hanson Group for the seven-figure costs of external lab testing and an early bulk purchase of more than 100,000 caps—that happened just last year. But the bustling scene suggests that business is good today. In the warehouse, towering stacks of Guardian Caps inventory have overtaken what was once an employee exercise area, like cubed cardboard weeds. In the loading dock, a truck drops off a fresh supply of green and maroon caps air-shipped from the company’s factory in Dongguan, China after their stockpile of those colors ran out.

When it comes to the Guardian Cap, the XTs sell for about $70 at retail stores and online but less than $55 in group orders for teams, while the NXTs are only sold direct to teams at an average of $100 apiece. According to the company, some 77 percent of its overall sales happens in bulk. But only a small portion of that total—about 200,000 caps sold this year, the company projects—comes from the NFL’s 32 teams, each of which typically stocks about 100 per season, replacing them every one to two years depending on the position and wear.

Not counting the cost of NFL testing through Biocore, though, Erin says Guardian has “never received a dime” directly from the league. Adds Lee, “They own the data too—only when they publish it, we get to see it.” But the benefit of having a public stamp of approval in the shape of the NFL shield is obvious.

“We had to have outside validation, and that’s what the NFL has done for us,” Erin says. “You can’t put a value on the marketing that you get from your product. But they’re not writing us checks, that’s for sure.”

On the flip side, the company has already done more for the NFL than its owners ever imagined. “The last thing we ever wanted to be in was the customization business,” Erin says, referring to the nine months it took to create the Guardian Cap pinnies that matched the helmets of every team so they could be worn in games this season. Between designing the concepts, conducting testing through Biocore, enlisting two college teams (Colorado and Georgia) to pilot them at spring practice, and ensuring the proper pantones and logos—all but four sent the initial batch back, including the Carolina Panthers, who requested that the silver pinnie “sparkle more,” Erin says—every step was geared toward aesthetic goals. Asked if it was a necessary headache to collaborate with the league, Lee replies immediately: “Yes.”

Adds Erin, “If it allows a player who feels like he's getting the same benefit in a game as in practice, then it's worth it.”

To this end, Guardian recently completed production on versions 2.0 of both the XT and NXT caps, featuring flatter, more helmetlike exteriors onto which teams can iron on logo decals. “The barriers to gameplay, I think, are the lack of gameplay testing data and the aesthetics,” says chief operating officer Jake Hanson, Erin and Lee’s son. “The more we can get at those with the 2.0s, the better.”

Given the relative lack of caps in NFL stadiums so far this season, it seems unlikely that they can catch on as a widespread gameday add-on. But even their future in practice remains uncertain. “The Guardian Cap could be a transition as we move from good helmets to better helmets, as that technology continues to improve,” Miller says.

That time might already be here: Entering this season, Biocore approved six new helmet models that would allow any player who wore one to be exempt from the Guardian Cap practice mandate. And more than 200 players have already taken advantage of the carve-out, “an unprecedented rate of adoption for a new piece of equipment,” NFL chief medical officer Allen Sills said on a media call in early October.

To both the lab and the league, the decision here boiled down to incentivizing players to use better-performing helmets.

“Decreasing benefit to wearing it, to the point where some of the technology that the helmets have now made the addition of a Guardian Cap on top of it of less utility than it would’ve been with one of the other helmets,” Miller says, when asked why the coverings were made optional at all. “And eventually you get to the plus-minus of, well, you’re adding a bunch of ounces to your head.”

Should the Hansons buck the odds once more and become a bigger NFL game-day presence down the line, it might be former players turned coaches leading the way. Like Manning. Or former running back and fellow Hall of Famer Jerome Bettis, an early proponent who bought Guardian Caps for an Atlanta-area youth football league that he and ex-Pittsburgh Steelers teammate Tim Lester started for their sons in the mid-2010s.

“You'll get these comments about, how can you put these things in the game?” Erin says. “Then you have interactions with people who are actually living the repercussions, it's a whole different thing.”

As for one of the few current players to have embraced the Guardian Caps in games, the Colts’ Granson doesn’t see himself stopping anytime soon.

“A big hit’s gonna happen—that’s the inevitability of the game,” he says. “But if I can reduce the contact of the little hits, then that’ll spare me down the road.”

The code hosting giant is expanding Copilot beyond its OpenAI-only architecture to include Anthropic’s Claude 3.5 Sonnet and Google’s Gemini 1.5 Pro models while simultaneously launching GitHub Spark to simplify web application development through natural language instructions.