YouTube Partner Program (YPP) 2025: Current Eligibility Requirements

YouTube now offers two tiers within the Partner Program:

1. Expanded YouTube Partner Program (Early Monetization Access)

• Subscribers: 500
• Public Uploads (last 90 days): 3
• Watch Hours (last 12 months): 3,000 OR Shorts Views (last 90 days): 3 million
• Monetization Features: Fan Funding (Channel Memberships, Super Chat, Super Stickers, Super Thanks) and YouTube Shopping access

2. Full YouTube Partner Program (Ad Revenue Sharing and Full Benefits)

• Subscribers: 1,000
• Watch Hours (last 12 months): 4,000 OR Shorts Views (last 90 days): 10 million
• Monetization Features: All features, including ad revenue sharing, YouTube Premium revenue, Fan Funding, Shopping, and BrandConnect

Additional Requirements for All Tiers:

• No active Community Guidelines strikes
• Adherence to YouTube’s Monetization Policies
• 2-Step Verification enabled
• Active AdSense account linked

Clarifying the Confusion: “3 Valid Public Watch Hours”

There is no YouTube requirement for “3 valid public watch hours” for monetization.

The confusion likely stems from two actual criteria:

• “3 valid public uploads” in the last 90 days (for expanded YPP eligibility)
• 3,000 valid public watch hours in the past 12 months (expanded YPP threshold)

In reality, creators must aim for thousands of valid watch hours, not just a few, to qualify for any monetization benefits.

What Are “Valid Public Watch Hours”?

“Valid public watch hours” are the total time viewers spend watching a channel’s public long-form videos.
Here’s what counts and what doesn’t:

Counted:

• Publicly available long-form videos
• Public livestreams converted into VOD (Video On Demand)

Not Counted:

• Private or unlisted videos
• Deleted videos
• Watch hours from paid ad campaigns
• YouTube Shorts views in the Shorts Feed (separate criteria apply)

Importantly, watch hours operate on a rolling 12-month basis, so older watch time eventually falls off if new engagement isn’t sustained.

Understanding Watch Hours vs. Earnings

While meeting the watch hour and subscriber requirements makes a channel eligible for monetization, it does not guarantee income.

Earnings are influenced by:

• RPM (Revenue per Mille): Revenue per 1,000 ad impressions
• CPM (Cost per Mille): What advertisers pay for 1,000 impressions
• Audience demographics and geography
• Ad engagement rates
• Video length and ad placement

Additionally, features like Channel Memberships, Super Chat, and YouTube Shopping provide alternative revenue sources that depend more on direct audience engagement than ad views alone.

Monetization Features and Their Requirements

Here’s a quick look at feature access based on your YPP tier:

Key Misconceptions to Avoid

• Meeting thresholds does not guarantee monetization.
  Every application undergoes a human review for content quality and policy adherence.
• Watch hours in the “Analytics” vs “Earn” section can differ.
  Only the “Earn” section shows eligible monetization watch time.
• Copyright violations or reused content can prevent or revoke monetization.
  Originality and authenticity are crucial for long-term success.
• Monetization application delays are common.
  Patience and adherence to platform guidelines are key.

Conclusion

In summary, the claim of needing “3 valid public watch hours” for YouTube monetization is incorrect. As of 2025, creators must meet either 3,000 or 4,000 valid public watch hours (depending on the YPP tier) or achieve high Shorts view counts to monetize.

Focusing on engaging, original, policy-compliant content and understanding the nuances of YouTube’s monetization process will set creators on a successful path toward sustainable earnings.

Always refer to the official YouTube Help Center for the latest, accurate information on eligibility requirements and monetization updates.

The post Understanding YouTube Earnings: Clarifying Watch Hour Requirements (2025 Update) appeared first on Technology Hobby.

Introduction

In recent years, advancements in artificial intelligence have paved the way for a new breed of digital assistants—AI agents—capable of executing complex tasks autonomously. These advanced systems differ significantly from traditional chatbots. They can automate repetitive tasks and possess reasoning capabilities that enable them to plan, decide, and act on your behalf.

At the same time, while many technologists and industry leaders are optimistic about the transformative power of agentic AI, a growing chorus of experts, as highlighted in sources like The New York Times and New Scientist, remain skeptical that current models will lead to accurate artificial general intelligence (AGI). In this blog, we delve into how AI agents are reshaping workflows and examine the debate around AGI, blending practical benefits with critical reflections on the future of AI.

What Are AI Agents?

Defining AI Agents

AI agents are advanced digital assistants that combine machine learning with autonomous decision-making. Unlike traditional chatbots that simply generate responses based on pre-programmed scripts, AI agents can:

• Reason and plan: They break down complex tasks into actionable steps.
• Automate tasks: They handle everything from scheduling appointments to processing data.
• Learn and adapt: They continuously improve performance based on new data and feedback.

These characteristics enable AI agents to optimize workflows, reduce human error, and ultimately drive significant efficiency gains.

Core Capabilities and Differentiators

While many digital tools exist for automation, what sets agentic AI apart is its ability to think beyond rigid rules:

• Autonomous Decision-Making: AI agents analyze situations, make informed decisions, and execute them without constant human oversight.
• Contextual Understanding: They leverage deep learning models to interpret complex contexts, making them invaluable in dynamic environments.
• Integration with Existing Systems: Seamlessly integrating with enterprise software, AI agents enhance workflows by bridging gaps between disparate systems.

The Evolution of AI in the Workplace

From Early Experiments to Advanced Systems

AI’s journey from rudimentary algorithms to today’s sophisticated models has been remarkable. Early experiments in machine learning laid the groundwork, but recent breakthroughs in deep learning and transformer models have unlocked capabilities once thought to be the realm of science fiction. Milestones, such as the development of large language models, have enabled AI to handle simple tasks and complex problem-solving.

Key Breakthroughs and Technological Milestones

• Deep Learning Revolution: The emergence of deep learning allowed machines to process and learn from vast amounts of data.
• Transformer Models: Innovations like GPT-4 and its successors have paved the way for AI agents that can understand and generate human-like text.
• Autonomous Reasoning: Today’s models incorporate advanced reasoning capabilities, enabling them to plan multi-step tasks—a critical step toward true agentic AI.

How AI Agents Transform Workflows

Enhancing Productivity and Automating Tasks

AI agents are proving to be game changers in various industries by streamlining workflows and automating repetitive tasks. Imagine an office where mundane administrative tasks—scheduling meetings, sorting emails, or even drafting routine reports—are handled by an AI assistant, freeing up human employees for more strategic work.

Real‑World Applications

• Finance: AI agents process financial data, generate reports, and even assist in fraud detection.
• Healthcare: They schedule patient appointments, manage electronic health records, and support diagnostic processes.
• Tech Startups: Agile startups deploy AI agents to streamline coding, testing, and data analysis, significantly reducing time-to-market.

Case Study: Productivity Gains

Consider a company implementing an AI agent to handle customer queries and internal documentation. After integration, the company saw a 30% increase in overall productivity and a noticeable reduction in error rates, allowing employees to focus on innovation rather than routine tasks.

Benefits of Embracing Agentic AI in the Workplace

Efficiency and Cost Savings

One of the most compelling reasons for businesses to adopt AI agents is the potential for significant efficiency gains. By automating routine tasks, companies can reduce labor costs and reallocate resources to more value-added activities. The cost savings are not just financial—they also translate to time saved, which can be reinvested into strategic initiatives and innovation.

Innovation and Competitive Advantage

Businesses that leverage agentic AI often find themselves at a competitive advantage. Companies can respond more quickly to market changes by streamlining workflows and reducing operational bottlenecks. Moreover, the insights generated by AI agents can lead to better decision-making and open up new avenues for growth and innovation.

Enhanced User Experience

From an employee standpoint, AI agents reduce workload stress and allow for more meaningful work. For customers, faster response times and more personalized services enhance overall satisfaction and loyalty.

The Debate on AGI: Optimism vs. Skepticism

Emerging Perspectives on AGI

There is growing excitement and apprehension about the advent of artificial general intelligence (AGI). Articles like The New York Times’s “Why I’m Feeling the AGI“ capture the emotional rollercoaster of anticipating a future where machines might match or exceed human intelligence one day.

Scientific Skepticism

However, while some see these advances as the first steps toward AGI, many AI scientists are skeptical. A recent New Scientist article highlights that modern deep learning models—despite their impressive capabilities—may not naturally evolve into true AGI. The main arguments include:

• Pattern-Matching Limitations: Current models excel at pattern recognition but lack genuine understanding or consciousness.
• Architectural Constraints: Existing architectures might require revolutionary breakthroughs to achieve the kind of general intelligence seen in humans.
• Data Dependency: AI agents rely heavily on vast datasets, and while this drives performance, it may also limit the models’ ability to reason abstractly or creatively.

Balancing Optimism with Reality

While businesses are already reaping the benefits of agentic AI, it’s essential to maintain a balanced view. True AGI—if it ever arrives—may require novel approaches beyond the current paradigm. The focus remains on practical applications that drive efficiency and innovation.

Challenges and Considerations in AI Adoption

Integration Hurdles

Integrating AI agents into existing workflows is not without challenges. Compatibility issues with legacy systems, the need for specialized training, and potential disruptions during the transition phase are all real concerns. Companies must invest in technology and change management strategies for a smooth transition.

Ethical and Privacy Concerns

With increased automation comes increased responsibility. Deploying AI agents raises important ethical questions regarding data privacy, transparency, and accountability. Organizations must establish robust governance frameworks to manage these risks and maintain trust among employees and customers.

Workforce Implications

The impact of AI on the workforce is a subject of intense debate. While AI agents can free employees from mundane tasks, job displacement is also risky. Upskilling and reskilling programs will help workers adapt to new roles and responsibilities in an AI-driven world.

The Future of Work: Letting AI Take the Wheel

Future Predictions and Business Strategy

Looking ahead, the future of work is set to be dramatically transformed by AI agents. As these digital assistants become more sophisticated, they will handle routine tasks and support strategic decision-making. Businesses that adopt AI solutions early will likely gain a significant competitive advantage.

Practical Advice for Businesses

• Start Small: Begin by automating repetitive tasks and gradually expanding the scope of AI integration.
• Invest in Training: Ensure your workforce is equipped with the skills to work alongside AI agents.
• Monitor and Iterate: Continuously assess the performance of AI solutions and refine them based on feedback and evolving business needs.
• Ethical Guidelines: Develop and enforce ethical guidelines to ensure transparency and accountability in AI usage.

Embracing the AI Revolution

By adopting AI agents, companies can create smarter, more efficient workflows that pave the way for future innovation. While the path to true AGI remains uncertain, the practical benefits of agentic AI are clear—streamlined operations, cost savings, and enhanced user experiences are just the beginning.

Conclusion

The transformative power of AI agents is already evident in today’s workplaces. These digital assistants are revolutionizing workflows by automating tasks, boosting productivity, and driving innovation. However, as we embrace the benefits of agentic AI, it is also important to recognize the ongoing debate around AGI. While some herald the dawn of a new era of machine intelligence, many experts remain skeptical that modern deep learning models will naturally evolve into true AGI.

For businesses, balancing ambition with caution is leveraging AI agents to optimize workflows and improve efficiency while maintaining ethical standards and investing in workforce training. The future of work is unfolding before our eyes, and the companies that adapt now will be best positioned to thrive in an increasingly digital and automated landscape.

Join the conversation: How are you integrating AI into your workflows? What challenges have you faced, and how do you see the debate on AGI shaping the future of your industry? Share your thoughts in the comments below.

FAQs

What is an AI agent?

An AI agent is a digital assistant that uses advanced machine learning and reasoning capabilities to perform tasks, make decisions, and streamline workflows autonomously.

How do AI agents improve workflow efficiency?

By automating repetitive and complex tasks, AI agents reduce manual work, minimize errors, and allow employees to focus on strategic, high-value activities.

Why are some experts skeptical about modern AI models leading to AGI?

Many AI scientists argue that while current models excel at pattern recognition, they lack true understanding and require fundamentally new architectures or breakthroughs to achieve genuine AGI.

The post How AI Agents Are Transforming Workflows and What It Means for AGI: Reality vs. Skepticism appeared first on Technology Hobby.

Despite employers requiring their employees to complete yearly cybersecurity training courses, human-driven cybersecurity breaches still happen. The problem could even get substantially worse as generative AI increases the scale and personalization of social engineering campaigns.

Anagram, formerly known as Cipher, is taking a new approach to employee cybersecurity training that the company hopes can keep up with the changing nature of these campaigns.

The New York-based company built a platform that contains hands-on security training for enterprises. The training includes bite-sized videos and personalized interactive puzzles to teach employees how to spot suspicious emails and communication. These trainings are designed to be more frequent, and more engaging, than the current standard of a once yearly, lengthy training session.

Harley Sugarman, the founder and CEO of Anagram, told TechCrunch that these activities include tasks like having employees create their own personalized phishing emails to teach them how to spot sophisticated campaigns against themselves.

“We took very little, in fact, basically no inspiration from the existing stuff out there,” Sugarman said regarding existing cybersecurity training. “What we really took was lessons from TikTok, and lessons from Duolingo and Khan Academy. We looked at these platforms that have done really, really well engaging and changing user behavior outside of the security space and we said, OK, how can we apply those lessons within security?”

Building gamified cybersecurity training wasn’t what Sugarman, a former VC at Bloomberg Beta, set out to do when he initially launched the company.

Sugarman’s first idea was a way to take the cybersecurity industry’s “capture the flag” training approach to upskill enterprise cybersecurity employees. This training approach involves building software with vulnerabilities and having security researchers go into the software to find the bugs and figure out how to write code without falling into the same traps.

That company launched as Cipher in 2022 and gained some traction. But chief information security officers (CISOs) started telling Sugarman that their businesses actually had a bigger security issue they were looking to tackle: their non-security employees. Sugarman said that CISOs describe their employees as their weakest cybersecurity link.

“What sort of surprised me was actually just the amount of hopelessness that I heard in their voices,” Sugarman said. “This was an unsolvable problem for them.”

Cipher then pivoted in January 2024 to focus on solving that problem. Now the startup is changing its name to Anagram to reflect its new focus and is in the process of winding down its original product. Anagram has seen strong growth since its pivot and landed customers including Thomson Reuters, MassMutual, and Disney, among others.

Anagram recently raised a $10 million Series A round led by Madrona with participation from General Catalyst, Bloomberg Beta, and Operator Partners, among others. The company plans to use the funds to build out its sales team and continue to improve the product. Sugarman said that so far they have been able to bring company’s phishing failure rates from 20% down to 6%, but he thinks they can continue to get closer to zero.

Sugarman said Anagram launched its product at a really interesting inflection point for the cybersecurity industry. With the advancements of generative AI, social engineering campaigns can be more personalized than ever, which will make it increasingly hard for people to tell what is real and what isn’t.

“I think the sort of side effect of that is that traditional email security platforms are actually going to have a much harder time detecting these AI-generated phishes,” Sugarman said. “That ability to generate and randomize is just so strong, and it’s really, really difficult, from an engineering perspective, to defend against that.”

Anagram is also working to develop an AI agent that will sit in enterprise employees’ emails and will be trained to flag potential cybersecurity slip-ups before they happen. Sugarman said the agent would do things like pop up to ask someone if they really want to send their credit card information over email and other similar safeguards.

In the meantime, Anagram hopes its puzzles and TikTok-like training videos will continue to move the needle.

“Humans are not dumb, we built skyscrapers we can do space travel,” Sugarman said. “We can figure out how to not click on a suspicious link in an email.”

The post Anagram takes a gamified approach to employee cybersecurity training appeared first on Technology Hobby.

Since its founding in 2021, OpenAI Startup Fund has raised $175 million for its main fund and secured an additional $114 million through five separate special purpose vehicles, which are investment pools for specific opportunities.

Unlike many sizable tech companies, OpenAI says it doesn’t use the company’s money to invest in startups.  The ChatGPT maker says its OpenAI Startup Fund is raised from outside investors. This includes participation from significant OpenAI backer Microsoft, as well as “other OpenAI partners,” according to the fund’s website.

The OpenAI Startup Fund, which is managed by a dedicated team, has so far invested in over a dozen startups, according to data providers PitchBook and Crunchbase and TechCrunch’s research. 

The following companies, organized alphabetically, have themselves announced investments from the fund.

1X: This Norwegian humanoid robot startup raised $23.5 million in a deal led by the OpenAI Startup Fund and Tiger Global in early 2023. However, OpenAI’s fund wasn’t named as a participant in the company when it announced its $100 million Series B in January.  

Ambience Healthcare: This AI-powered medical note-taking startup announced a $70 million Series B in February 2024, co-led by OpenAI’s fund and Kleiner Perkins. Ambience is among a number of startups, including Abridge, Nabla, Suki, and Microsoft-owned Nuance, that are building AI medical scribes.

Anysphere (aka Cursor): In October 2023, OpenAI’s fund led the $8 million seed round into Anysphere, the maker of AI-powered coding assistant Cursor. OpenAI hasn’t been named as an investor in the company’s subsequent rounds.

Chai Discovery: This startup, which is developing an open source AI foundational model for drug discovery, raised a $30 million in seed round led by Thrive Capital and OpenAI’s fund last September. The deal valued the 6-month-old Chai Discovery at $150 million.  

Class Companion: This edtech startup raised a $4 million seed round in 2023 from OpenAI’s fund and a host of angels. It helps teachers provide quick, personalized feedback to their students.  

Descript: The collaborative audio and video editing platform raised $50 million in a Series C round led by OpenAI’s fund shortly after ChatGPT was introduced to the world in late 2022. Other investors in the round included Andreessen Horowitz, Redpoint Ventures, Spark Capital, and ex-Y Combinator partner Daniel Gross. Descript hasn’t reported any other capital raises since its Series C.

Figure AI: AI robotics startup Figure raised a $675 million Series B in February 2024 from Nvidia, OpenAI’s fund, Microsoft, and others. The round valued the company at $2.6 billion. Figure AI is now reportedly in talks to raise $1.5 billion at a $39.5 billion valuation.

Ghost Autonomy: This maker of autonomous driving software raised a $55 million Series E in April 2023, and OpenAI’s fund invested $5 million of that, according to PitchBook data. But the investment didn’t work out. A year later the company shut down.

Harvey AI: This legal tech startup raised a $21 million Series A in April 2023 from OpenAI’s fund and others. The fund also participated in three subsequent rounds, including last month’s $300 million Series D, which valued Harvey at $3 billion.

Heeyo: Educational AI chatbot for kids Heeyo announced that it raised $3.5 million from OpenAI’s fund, Alexa Fund, Pear VC, and other investors in August.

Kick: This company is developing AI agents that it says can “self-drive” bookkeeping processes. It raised a $9 million seed round co-led by General Catalyst and OpenAI’s fund in October.

Mem: This AI-powered note-taking startup raised a $23.5 million Series A round in November 2022, led by OpenAI’s fund. Mem hasn’t reported any subsequent funding rounds.  

Milo: This startup is developing an AI-powered personal assistant that helps parents organize and keep track of their kids’ activities. Milo raised an undisclosed amount of pre-seed and seed funding from OpenAI’s fund, YC, and others.

Physical Intelligence: Foundational software for robots startup Physical Intelligence raised a $70 million seed round last March. OpenAI’s fund was part of that and also participated in the company’s $400 million Series A that valued the company at more than $2 billion. Other investors in the latest round included Lux Capital, Sequoia, and Jeff Bezos.

Speak: This AI-powered language learning app developer raised a $27 million Series B round in November 2022, led by OpenAI’s fund. In December, the fund participated in Speak’s $78 million Series C, which valued the company at $1 billion.

Thrive AI: Huffington Post founder Arianna Huffington and OpenAI Startup Fund announced last July that they teamed up on investing and building this “AI health coach” startup. Thrive AI aimed to raise $10 million, according to a regulatory filing.

Unify: This sales technology startup secured about $19 million in seed and Series A capital from the OpenAI Startup Fund, Thrive Capital, and Emergence.

The post OpenAI’s startup empire: The companies backed by its venture fund appeared first on Technology Hobby.

TC Sessions: AI is an event like no other that will let you explore the AI industry with a focus on the startup ecosystem. Whether you’re a founder or an investor or you’re simply curious about the wild world of AI, join us on June 5 in Zellerbach Hall at UC Berkeley to get inspired and equipped for your next big idea.

Secure your spot before March 2 at 11:59 p.m. PT to save at least $300.

Why attend TC Sessions: AI? 

At TC Sessions: AI, you can gain invaluable insights alongside AI leaders, VCs, and enthusiasts. The jam-packed day features breakout sessions with the best minds in AI, hands-on product demos, and panels from AI trailblazers.

Launching an AI product against established competitors

How do you stay relevant against big players in a highly competitive market? Odyssey co-founder and CEO Oliver Cameron’s panel at TechCrunch Sessions: AI will look at ways small companies can compete against their rivals in a fast-paced and rapidly changing space. 

Cameron is looking at the next frontier of artificial intelligence by pioneering “world models.” His startup Odyssey is training an AI model that aims to generate “cinematic, interactive worlds in real time.” Cameron previously was the co-founder and CEO of autonomous vehicle startup Voyage.

TC Sessions: AI will also include insightful conversations from Twelve Labs CEO Jae Lee, CapitalG partner Jill Chase, and Khosla Ventures partner Kanu Gulati. To learn more about who else will join us on the main stage, check out the TC Sessions: AI event page or apply here by March 7 for a chance to be a speaker yourself.

Don’t miss out on these unbeatable prices

TC Sessions: AI is the place to learn and network. Looking to pitch to investors or bounce ideas off others in intimate group discussions? This is the time — and place — to do it.

Ready to learn from the AI experts? Register now to save up to $325 on select tickets. But don’t wait — this special deal ends on March 2 at 11:59 p.m. PT. 

Want more deals and promotions to TechCrunch events like this? Subscribe to our TechCrunch Events newsletter and you’ll be the first to know when they happen.

Make your voice heard at TechCrunch Sessions: AI

Are you an AI expert who can drive powerful discussions with tech innovators and entrepreneurs? We want to hear from you! Apply by March 7 for a chance to share your expertise and shape the future of the AI community.

Is your company interested in sponsoring or exhibiting at TechCrunch Sessions: AI? Contact our sponsorship sales team by filling out this form.

The post 2 days left to save up to $325 on TechCrunch Sessions: AI tickets appeared first on Technology Hobby.

Chinese AI startup DeepSeek recently declared that its AI models could be very profitable — with some asterisks.

In a post on X, DeepSeek boasted that its online services have a “cost profit margin” of 545%. However, that margin is calculated based on “theoretical income.”

It discussed these numbers in more detail at the end of a longer GitHub post outlining its approach to achieving “higher throughput and lower latency.” The company wrote that when it looks at usage of its V3 and R1 models during a 24-hour period, if that usage had all been billed using R1 pricing, DeepSeek would already have $562,027 in daily revenue. 

Meanwhile, the cost of leasing the necessary GPUs (graphics processing units) would have been just $87,072.

The company admitted that its actual revenue is “substantially lower” for a variety of reasons, like nighttime discounts, lower pricing for V3, and the fact that “only a subset of services are monetized,” with web and app accessing remaining free.

Of course, if the app and website weren’t free, and if other discounts weren’t available, usage would presumably be much lower. So these calculations seem to be highly speculative — more a gesture towards potential future profit margins than a real snapshot of DeepSeek’s bottom line right now.

But the company is sharing these numbers amidst broader debates about AI’s cost and potential profitability. DeepSeek leapt into the spotlight in January, with a new model that supposedly matched OpenAI’s o1 on certain benchmarks, despite being developed at a much lower cost, and in the face of U.S. trade restrictions that prevent Chinese companies from accessing the most powerful chips. Tech stocks tumbled and analysts raised questions about AI spending.

DeepSeek’s tech didn’t just rattle Wall Street. Its app briefly displaced OpenAI’s ChatGPT at the top of Apple’s App Store — though it’s subsequently fallen off the general rankings and is currently ranked #6 in productivity, behind ChatGPT, Grok, and Google Gemini.

The post DeepSeek claims ‘theoretical’ profit margins of 545% appeared first on Technology Hobby.