Through OpenAI's internal whistleblowing, lack of transparency, and astronomical debt, we examine the current situation of this massive AI firm and what we should be paying attention to.
AI is changing the world, but is the company behind it shaking?
Imagine this: What if the AI assistant or chatbot that helps you with your work every morning were actually a ‘black box’ with an interior as hollow and unknowable as an empty box? And what if the company running that box were sitting on a massive mountain of debt?
When we think of ‘Artificial Intelligence (AI),’ we often imagine cutting-edge technology and a dazzling future. However, inside the massive companies bringing that technology to the world—especially OpenAI, famous as the developer of ‘ChatGPT’—some unsettling things are happening right now. Today, I want to step away from complex technical jargon and walk through the conflicts and crises happening behind the scenes of the AI services we use daily.
Why does this matter?
AI has become a part of our lives. When students do homework or office workers write reports, we subconsciously turn to AI for answers. The reason we trust and use AI is because of the expectation that it will be ‘safe’ and ‘smart.’
However, the noise surrounding OpenAI raises important questions for us: “Is the AI we use truly safe?” and “Are they putting our safety aside to make money?” This issue is not just a concern for a single company, but a critical point of debate that will set the standards for how we trust and accept AI technology moving forward.
Simple Understanding: ‘Black Box’ and ‘Safe Dog’
To understand OpenAI’s situation, let’s look at two analogies.
First is the ‘black box.’ When GPT-4 first came out, OpenAI disclosed nothing about the model’s size, what data it had studied, or how it was built. Source: GPT-4’s Secret Has Been Revealed - by Alberto Romero It’s like eating a delicious meal while the chef refuses to tell you any of the ingredients. People call this a ‘black box.’ Since the contents are unknown, it’s only natural for users to sometimes feel anxious.
Second is the ‘safe dog.’ OpenAI promoted loudly on its website, “ChatGPT was built with safety in mind.” Source: Florida lawsuit accuses OpenAI of ignoring safety… - The Guardian In short, they marketed it as a well-trained, gentle dog, but when you’re out walking, the dog is actually threatening people. The recent lawsuit in Florida digs exactly into this point: the gap between marketing—claiming it’s safe—and the reality.
Even internal staff are raising issues. There is internal testimony that in the process of reviewing the safety of a certain AI model, the inspection was finished in just one week. Source: Unraveling OpenAI’s AI Safety Conundrum: Promises Unfulfilled. This feels just as dangerous as saying a highly complex major surgery was safely completed in just one minute.
Current Situation: Glamorous Surface and Hidden Truths
OpenAI currently faces a perplexing situation on multiple fronts. Beyond mere technical issues, financial problems and trust issues are intricately intertwined.
- Unimaginable Financial Pressure: The financial obligations OpenAI must resolve by the end of 2030 exceed a staggering $800 billion. Source: What Happens If OpenAI Dies? - Ed Zitron’s Where’s Your Ed At This is a number so huge that it is difficult for the average person to grasp. The costs to be paid to large cloud companies like Oracle, Google, Amazon, and Microsoft alone are estimated to reach $146 billion by 2027.
-
Insiders’ Voices: Former colleagues and current employees are publicly expressing concerns about the company’s financial motives and its responsible AI operational practices. [Source: It’s all unraveling at OpenAI (again) Business Insider via Yahoo News](https://ca.news.yahoo.com/unraveling-openai-again-214510558.html) -
Endless Conflict: Disputes and noise are relentless, including lawsuits with Elon Musk, conflict with actress Scarlett Johansson, lack of transparency controversies, and strict non-disclosure agreements (NDAs). [Source: It’s all unraveling at OpenAI (again) Yahoo Tech](https://tech.yahoo.com/ai/articles/unraveling-openai-again-214510558.html)
What will happen next?
The ‘tangled knot’ of a situation OpenAI is experiencing is likely to continue for some time. However, one thing is clear: there is a limit to how much ‘black box-style operation’ can persist if a company wants to earn public trust in the AI era.
To use an analogy, OpenAI now stands at a crossroads where it must decide whether to become a ‘transparent glass box’ or hide further into the darkening shadows. What we should pay attention to moving forward is how transparently OpenAI discloses information and how effectively it strengthens its safety verification procedures. Additionally, how they will manage massive debt while continuing to provide AI services we can use with peace of mind will be a key point to watch.
## References
- GPT-4’s Secret Has Been Revealed - by Alberto Romero
- Florida lawsuit accuses OpenAI of ignoring safety… - The Guardian
- Unraveling OpenAI’s AI Safety Conundrum: Promises Unfulfilled.
- What Happens If OpenAI Dies? - Ed Zitron’s Where’s Your Ed At
-
[It’s all unraveling at OpenAI (again) Business Insider via Yahoo News](https://ca.news.yahoo.com/unraveling-openai-again-214510558.html) -
[It’s all unraveling at OpenAI (again) Yahoo Tech](https://tech.yahoo.com/ai/articles/unraveling-openai-again-214510558.html)
- Lack of financial transparency
- The gap between marketing emphasizing safety and actual operations
- Conflict with Elon Musk
- $45 million
- $90 million
- Over $800 billion
- Because its internal structure and training data were not disclosed
- Because it consumes too much power
- Because anyone can modify it