Service Soars, Penske Fills The Best-list, Is Chat GPT Getting Dumber?

Shoot us a Text.

As many kids start school, there is still plenty of summer to go as we talk about soaring service department numbers. We also talk about Penske taking an early victory lap regarding the best places to work list, as well as an interesting development in ChatGPT's ability to do math. 

Show Notes with links:

• With consumers keeping their cars longer due to affordability and availability issues, the six largest publicly traded groups are reporting an 8.2% increase in YoY sales logging $3.9 B in revenue. 
  • AutoNation reports a 9.1% first half increase of same store sales as CEO Mike Manley credits a large existing customer base who know the AutoNation brand, extensive service bay and technician capacity and customers keeping and servicing their vehicles longer.
  • Lithia is up 5.8% for the quarter and COO Chris Holzshu made similar comments in a recent investor call, 
    • “With the average age of vehicles rising, coupled with the increasing complexity of new vehicles, our team of certified, factory-trained technicians are working hard to deliver on the massive demand for services from our customers,” he says.
  • Penske 10.7,  Sonic 8.9, Asbury 6.2, Group 1 reported 9 consecutive quarters of growth
• Penske Automotive Group is taking an early victory lap as they announce a significant achievement as 44 of its dealerships are named among the Best Dealerships to Work for in 2023 by Automotive News. 
  • The annual list, now in its 12th year, recognizes top dealerships nationwide for their exceptional work environments and community engagement
  • For inclusion on the list, dealerships undergo a thorough nomination procedure. Selection of winners hinges on feedback from employee surveys, active community participation, and deliberate actions to convey gratitude and acknowledgment in the workplace.
  • The full list will be released in October.
• Is ChatGPT getting dumber? This is a question asked in a WSJ article as the groundbreaking language model has become worse at performing certain basic math operations, according to new research by Stanford University and UC Berkeley.
  • The research tested two versions of ChatGPT, 3.5 and 4.0, and found that the latter's accuracy in identifying prime numbers dropped from 84% to 51% between March and June.
  • Six out of eight tasks saw worsening performance in the newer version, while the older version showed improvement in six areas but still remained less capable in most tasks.
  • The article implies that "prompt engineering" can induce drift in AI models. In attempts to prevent inappropriate AI responses, developers may inadvertently impair the AI's capabilities in areas like math problem-solving.

Join Paul J Daly and Kyle Mountsier every morning for the Automotive State of the Union podcast as they connect the dots across car dealerships, retail trends, emerging tech like AI, and cultural shifts—bringing clarity, speed, and people-first insight to automotive leaders navigating a rapidly changing industry.

Get the Daily Push Back email at https://www.asotu.com/

JOIN the conversation on LinkedIn at: https://www.linkedin.com/company/asotu/