What Did OpenAI Do This Week? - 28/05/2023 [+20 LINKS]

OPENAI SAYS, “JK, THANK EU”

A busy week for OpenAI. The company announced a new grant program and massively expanded the reach of its iOS app while it toured those areas (the app was only launched in the US last week) while simultaneously threatening to leave the EU if the company can't comply with the EU's upcoming AI Act before later reversing the comment made during the whistle-stop tour of 17 counties in the EU.

Altman’s tour of the EU had gone without much controversy until this comment, which hit the headlines for a few reasons. One, it’s a massive business territory, important culturally, and any comment about leaving a massive territory is bound to raise eyebrows. The comment came surrounding Altman’s perception that the current draft of the EU AI Act was "over-regulating" just days after welcoming regulation in front of Congress.

The flipflop is important. It shows that OpenAI is talking a good game regarding regulation, but when the specifics are drawn, that regulation has to work in their favour (or at least not against it). Did Altman’s crown slip for a lot of people? Definitely (even with the backtracking). While that makes obvious business sense (what business wouldn’t want those conditions?), it’s hard to play the innovator, humanitarian, caring business luminary card when you threaten to leave if you don’t get your way.

Check out the links before for the full picture of everything that happening with OpenAI week.

SO WHAT?

So what’s the reality here? Could OpenAI pull out of Europe? Of course. Will it? No. Why? Money, culture, fairness, take your pick. Aside from that, any withdrawal really shows the bad side of any company as so far that the move shows a ‘we could play ball but won’t, and now you can all suffer’ attitude.

The bigger reality here? The EU doesn’t like allowing carte blanche approaches to companies with secret sauce (we don’t know what’s in GPT4) that disrupts industries and can impact lives. Why? They can’t assess risk, so they could harshly regulate to protect people and businesses. Some believe the argument is akin to asking Pepsi to reveal its recipe, but the difference in OpenAI’s case is the impact of the product, not giving away a competitive advantage. LLMs that OpenAI base products like ChatGPT affect more than those who use them and that level of impact and effect. Transparency is the key for OpenAI to allow regulation not to blow its lead successfully. The fact remains Altman et al can’t push around the EU or countries (see Italy) for a few reasons. OpenAI, while powerful, are tiny in comparison, and the EU holds the cards. OpenAI isn’t the only generative AI fish in the sea now; see Google, Meta, NVIDIA, and Amazon. Note that list doesn’t include any Chinese competitors, some of which are fast approaching the same levels as the US companies with a vastly greater test population. Additionally, while OpenAI and Microsoft are totally separate companies, Microsoft won’t want to upset the EU either for their business or the products that they are slamming OpenAI tech into. Doing so would make its life much harder if they had to suddenly strip them out or block any of the 749 million EU citizens.

WANT TO INNOVATE LIKE OPENAI? Order your copy of the second volume of ‘Disruptive Technologies’ now!