AI Top-of-Mind for 3.28.24 — Pairing Up
Top-of-mind is the continuing battle of the hyperscalers, now with AI added. Latest is Amazon’s added $2.5B investment in Anthropic. The ‘NY Times’ and Amazon provide more background and the fact that it includes access to the company’s AI systems and commitments for computing resources and use of Amazon technology.
On the job front, remember that the need for real creativity will never go away, and when some jobs disappear, new ones are created. But a view on entry and mid-level computer programming from Max Can’t Help It! Including the following insights:
• AI software, like GitHub Co-Pilot, will serve up code that has already been proven to work, by some anonymized developer (I wonder if some of my code is in its big brain), at whatever task you ask it. (What used to take 40% of my time).
• Co-Pilot can figure out what’s wrong with the code you’re developing — as you modify it to your specific purpose — when you post any gobbledygook error message you might encounter. (What use to take another 40% of my time).
• Even a single space or letter in the wrong place will render a program completely inoperable. AI doesn’t make those mistakes. Indeed, you can ask it questions with all manner of spelling and grammar problems — it usually figures out what you want and will deliver code free of typos.
• In short, Co-Pilot removes 80% of the specialized skills I once brought to the table.
Remember the OpenAI bot store announced last fall? But like with anything touched by the internet, it if filling with spam. Kyle Wiggers writing in ‘TechCrunch’ describes why, providing specific examples of copyright violations, tools to promote plagiarism (below), impersonations, and even bots designed to jailbreak OpenAI’s terms of use.
Turning to the equipment market, more strong growth signs on the back of AI. ‘Light Reading’ covers the latest IDC telecom survey including two good takeaways from Mario Morales, group VP for enabling technologies and chips:
• Morales said the number of connected edge devices will double in volume from 2022 to 2027 and ultimately will carry the bulk of AI traffic. Most of this will be AI inferencing, which is doing the data crunching that provides answers based on the existing models, rather than the training that creates the models.
• “AI inferencing will actually be larger than what we’ve already seen in training for within the data center and infrastructure, and we think that bodes well for the outlook of the market on a longer-term basis,” he said.
Data centers are also expected to grow dramatically, with 45% growth tipped for 2024. Over the next five years, IDC predicts data centers will become the biggest source of demand in the semiconductor market, overtaking PCs and smartphones.
Finally, some retail guidance, calling from a ‘customer-centric’ AI strategy. Scott Clark in ‘CMSWire’ reminds us to:
· Embrace AI innovation. A customer-centric AI strategy is vital for brands to personalize interactions and meet evolving customer expectations in the digital marketplace.
· Prioritize personalization. AI enables brands to deliver personalized experiences at scale, enhancing customer satisfaction and loyalty by anticipating and addressing individual needs.
· Invest in AI integration. Implementing a customer-centric AI strategy across all customer touchpoints is essential for creating seamless, personalized experiences and gaining a competitive edge.
