The company dropped two new open-weight reasoning models this week, gpt-oss-120b and gpt-oss-20b, which are available for free download on Hugging Face.

It’s OpenAI’s first open release since GPT-2 in 2019, marking a significant shift for a company that has built its business on proprietary, closed-source models.

The Real Story

Let’s be clear about what’s driving this decision. OpenAI didn’t suddenly develop a passion for open source. The company has been losing ground to Chinese AI labs like DeepSeek and Alibaba’s Qwen, which have dominated the open model space while OpenAI focused on charging enterprise customers for API access.

CEO Sam Altman acknowledged they’d been “on the wrong side of history” with their closed approach. That’s corporate speak for: we’re losing market share and need to adapt quickly.

The new models deliver solid performance. The 120B version outperforms DeepSeek’s R1 on coding benchmarks, while the lighter 20B model runs on consumer hardware with 16GB of memory. Both can connect to OpenAI’s cloud models when they encounter tasks beyond their capabilities — a strategic hybrid approach that keeps users within OpenAI’s ecosystem.

The Trade-offs

These models come with significant limitations. They hallucinate at rates of 49-53% compared to just 16% for OpenAI’s premium o1 model. OpenAI frames this as “expected” because smaller models have less world knowledge, but that’s still a substantial accuracy problem for commercial applications.

The models are also text-only, lacking the multimodal capabilities of OpenAI’s flagship products.

What This Means for Business

For entrepreneurs and developers, this represents a major opportunity. The Apache 2.0 license allows commercial use without royalties to OpenAI. However, the company isn’t releasing training data, likely due to ongoing copyright litigation.

The broader impact extends beyond just these models. This move forces competitors to accelerate their open source efforts, potentially triggering a new wave of accessible AI tools and increased competition across the sector.

Ready to understand how AI and emerging technologies are reshaping business strategies? Take Roger Hamilton’s Genius Generation microcourse for free at Genius Academy, and obtain insights on these technological shifts and how they not only change the way we do business, but also how we live.

The post OpenAI Releases First Open Models in Over 5 Years appeared first on Must-Read Blogs For Entrepreneurs | Wealth Dynamics.

The European Union just rolled out a new set of voluntary AI guidelines. Despite criticism from tech companies, we’d argue that it’s positive news on balance for entrepreneurs. 

Announced on July 10, these rules work with the EU’s AI Act (which kicked off in August 2024) to help companies like OpenAI, Google, Meta, and xAI manage risks from AI tools like Grok or ChatGPT. For business owners, this means a safer, more trustworthy AI landscape to grow your ventures.

The rules focus on being open about AI data, following EU copyright laws, and tackling risks that could affect people’s rights. Companies like France’s Mistral, which already signed on, get clearer legal paths and less paperwork hassle. 

For entrepreneurs using AI, whether for marketing, customer service, or operations, this is the pathway to a reliable environment. You can confidently use AI tools knowing they’re less likely to be misused, which builds trust with your customers.

The Computer and Communications Industry Association (CCIA), which includes tech behemoths like Amazon, Apple, and Google, say the rules are too strict, especially on safety and copyright. But for entrepreneurs, clear guidelines make it easier to pick trustworthy AI providers. You don’t have to worry about navigating murky regulations; the rules ensure AI tools are safe and ethical, letting you focus on growing your business with cool AI solutions.

The rules also require AI companies to share a summary of their plans before launching. This transparency helps you choose providers that align with your values, building confidence in the tools you use. Over 1,000 experts and community groups shaped these guidelines, balancing innovation with responsibility, while addressing key concerns like protecting rights.

For Wealth Dynamics entrepreneurs, these rules may be a golden opportunity. Using AI from compliant providers shows you’re ethical and forward-thinking, attracting customers and investors who care about doing things right. The focus on fairness and transparency fits perfectly with building sustainable businesses.

The EU’s approach sets a global standard, letting you leverage AI that’s safe, innovative, and built to help your business thrive.

The post EU’s AI Rules: A Boost for Ethical Entrepreneurs appeared first on Must-Read Blogs For Entrepreneurs | Wealth Dynamics.