- Tech Drop
- Posts
- The Tech Drop
AI companies aim for more engagement
The Drop: Instagram co-founder claims AI companies are designing their chat bots for user engagement instead of usefulness, while Open AI has had to roll back one of its newest models for being too friendly.
The details:
Kevin Systrom, the co-founder of instagram, said of AI companies “You can see some of these companies going down the rabbit hole that all the consumer companies have gone down in trying to juice engagement”.
Tactics such as asking additional questions and being overly friendly are being used to increase engagement metrics.
This comes as OpenAI has had to roll back it’s newest GPT-4o model due to complaints of it being overly “flattering”.
Why it matters: The future trend of AI companies seems to be moving in the same direction as social media companies, where capturing user attention and engagement is the ultimate metric that is maximised for.
Chip Wars
The Drop: Nvidia and Anthropic clash over chip exports, with Anthropic wanting tighter restrictions on chips going to China to maintain US dominance in AI
The details:
Anthropic released a blog post on April 30th calling for the US government to increase its restrictions on chip exports to China and implement these restrictions at an earlier date.
Nvidia criticised these comments stating “American firms should focus on innovation and rise to the challenge”.
Chinese companies such as DeepSeek acknowledge that chip restrictions into china are their main constraint when it comes to competing with US companies
Why it matters: AI competition will likely be one of the main flashpoints between the US and China in the years ahead. The US holds a major edge with access to advanced chips. However, restrictions can foster innovation, as has been seen with Deepseek, and China still has roughly half of all the world’s AI researchers.
AI demand outstrips supply
The Drop: Microsoft has admitted that demand for cloud based AI services is outstripping supply, this is despite huge recent investments in the area.
The details:
Amy Hood, Microsoft’s Chief Financial Officer, explained on the company’s latest earnings call that supply for AI cloud services on Azure has not kept up with demand.
Investment to increase capacity has increased 79% from last year to $14 billion in the first quarter of the year.
Hood said that capacity issues could impact results from AI and a supply issue means Microsoft can provide less AI services to cloud clients.
Why it matters: Companies are needing an increasing amount of computing power to run modern AI models. However, supply of advanced chips and the building of new data centers has lagged behind this jump in demand. In the short-term this could be a bottleneck for AI innovation and adoption.