Welcome to the AI Ecosystem Report, featuring practitioner analyst and entrepreneur Toni Witt. This series is intended to deliver the timely intelligence about artificial intelligence (AI) you need to get up to speed for an upcoming client engagement or board meeting.
This episode is sponsored by Acceleration Economy’s AI Ecosystem Course, available on demand. Discover how AI has created a new ecosystem of partnerships with a fresh spirit of customer-centric cocreation and renewed focus on reimagining what is possible.
Highlights
Innovation (00:51)
OpenAI’s DevDay, held on Monday, Nov. 6, was filled with innovative discussions. I recently shared some of my top takeaways from DevDay. One key highlight was that the CEO at the time, Sam Altman, acknowledged one area in which OpenAI is weak: allowing customers to build custom models on its infrastructure.
There’s much value in custom training, especially for business clients and enterprise use cases. For instance, it could help companies reach higher success rates, reduce hallucinations, and reduce risks. The organization claims it will build out customization offerings, but Altman didn’t elaborate on that.
Altman announced GPT-4 Turbo, a more powerful form of GPT-4. This model will enable users to propose longer, more complex prompts. Additionally, it will be cheaper than previous GPTs; this will lower the barrier to entry. The advancements made with GPT-4 Turbo will, hopefully, lead to much more innovation from more companies and startups.
Funding (04:43)
German-owned supermarket company Lidl is part of a $500 million investment in the German AI startup Aleph Alpha. While this startup is one of the only big AI companies in Europe, it only has around 70 employees. It aims to offer data protection, security, and transparency to clients.
Aleph Alpha has already raised about 30 million euros. In the United States, many generative AI startups are backed by platform partners that can help them go to market. However, in Europe, they’re often backed by the EU government or large enterprises that would be customers or users of the technology, hence Lidl’s investment.
Solution of the Week (7:00)
Carmax, one of the largest car retailers in the world, is leaning into AI with a new feature. The company collects online evaluations of a car 30 days after a customer buys it. It can be difficult to extract a lot of value from this unstructured data, especially from the perspective of a prospective buyer browsing the site.
The company’s search engine optimization (SEO) staff tasked GPT 3.5 to read through the evaluations for different models and makes of cars and then produce a summary paragraph that’s weighted to the evaluations. Of course, they do a human check of the final output before publishing it on the site. In natural language, the paragraph would reflect thousands of people’s reviews about the vehicle.
This project was spearheaded by Carmax CIO Shamim Mohammad. He pointed out that this initiative came from trial and error of the company’s SEO team, so it was a bottom-up effort. Mohammad also noted that other teams across the organization are tinkering with other generative AI applications.
