The Brief on AI #1
On the potential fall of OpenAI, Microsoft's big bet and the decline of ChatGPT
Happy Friday!
Here are 3 recent developments, 1 quote, my take on it all and 1 useful AI tip for you to apply or learn from.
3 recent developments
1. The New York Times and OpenAI could end up in court.
If OpenAI is found to have violated copyrights, the court could order the company to destroy ChatGPT's dataset and the company to face financial penalties of up to $150,000 per infringement. This could be a potentially fatal consequence for OpenAI.
2. Microsoft still stuck at 3% search market share
Microsoft invested 10 billion dollars in OpenAI hoping it would enable search engine Bing to surpass Google. It is still stuck at roughly 3%.
3. OpenAI traffic in decline since June
OpenAI grew to an astonishing 1900 million monthly visits in May '23, but saw a 300 million decline in June and lost another 100 million monthly visitors in July.
This week’s quote
"With every passing day, OpenAI looks more like Napster or the many defunct piracy platforms—it relies on the creativity of others to make a buck. And there are plenty of laws against that."
What to make of all of this?
We are seeing a classic example of the Gartner Hype Cycle in action. Every new promising technology experiences a phase of hype and inflated expectations. This is followed by disappointment because the expectations aren't met. The tech providers use this feedback to solve the problems and a new more permanent growth takes place. After the big dip and the recovery the real innovation starts to take place.
Source: Wikipedia
The potential law suit between OpenAI and The New York Times will be super interesting to follow but chances are significant they will find a way to collaborate because there is just too much money to be made. Even if they do end up in court and OpenAI faces the fatal consequences of copyright infringement this could mean the end of OpenAI, but not of LLMs.
Bing might not beat Google in search but users will increasingly rely on AI to provide them with specific answers, particularly when researching unfamiliar topics or completing specialized tasks such as coding and setting up SEO campaigns. So Microsoft might be losing on the short term, but on the long run chances are their strategy is going to pay off.
Useful AI tip
The tip of this week is provided by Ethan Mollick. A common way to get an LLM to explain something to you is by asking it to give an answer a 6 year old would understand. This doesn't always work so well. For some explanations the LLM should first ask questions to provide an explanation and Ethan figured out a prompt that does just that. Try it out here. You can run it in GPT-4 with this link (it will not work well in the free GPT-3.5, but can work in Bing in creative mode).
More context here:
