Amazon CEO shares his “thoughts on Generative AI”

In the message, Jassy reflected on what began as an ambitious dream for Amazon – to re-imagine the customer experience through AI – to now, where that dream is fast becoming a reality with the use of Generative AI: making the lives of its customers “better and easier” with this “once-in-a-lifetime” technology. 

According to Jassy, AI is enabling smarter, agent-driven assistants that can perform meaningful tasks – a leap forward from the AI-assisted chatbots that can only answer queries falling within a pre-defined set of data. 

He goes on to say that Amazon already has over 1,000 Generative AI tools in development or use, and it plans to create billions more AI agents that will work across businesses and daily life. These systems, he said, are delivering efficiency, personalisation, and performance across the consumer, seller, advertising, and Cloud divisions. 

The wagon-wheel effect: which way is the wheel spinning? 

Writing about positive outcomes of using AI and the plans for further rollout – such as AI agents “changing the way we work and live”, Jassy also highlighted that its advantage starts from the off: “Agents will allow us to start almost everything from a more advanced starting point. We’ll be able to focus less on rote work and more on thinking strategically about how to improve customer experiences and invent new ones.” 

He also emphasised the need to keep the company’s mentality like that of a startup: “we’re going to keep pushing to operate like the world’s largest startup – customer-obsessed, inventive, fast-moving, lean, scrappy, and full of missionaries trying to build something better for customers and a business that outlasts us all,” with “fun and exciting” being emphasised when it came to talking about how the technology will enhance the lives of everyone. 

However, this announcement did not come without a caveat. He went on to speak about the potential fallout from this technology: “We will need fewer people doing some of the jobs that are being done today, and more people doing other types of jobs,” he said, whilst noting that there is uncertainty about what this means for the future of its corporate workforce. “It's hard to know exactly where this nets out over time, but in the next few years, we expect that this will reduce our total corporate workforce as we get efficiency gains from using AI extensively across the company.” 

The wider fallout 

This message is a reflection of a wider industry trend, with an IMF study suggesting that roughly 60% of roles in advanced economies face potential disruption from AI. 

Dario Amodei, CEO of Anthropic, has even speculated in an interview with AXIOS that half of entry-level whitecollar roles could disappear, with US unemployment rising to as high as 20% in a worst case scenario. 

Geffory Hinton – widely regarded as the “Godfather of AI” – also errs on the side of caution when using the technology, warning that the dual threat of bad actors misusing AI and the existential risks posed by autonomous systems could overtake human control – something he never envisioned at the start of his career. In a 2023 MIT interview, Hinton warned that AI might soon surpass the human brain’s capacity and possibly become an intelligence beyond our control. 

Whereas other people, such as Susan L. Epstein, Professor of Computer Science, Hunter College and the Graduate Center, CUNY, has a different point of view, saying during a One to One interview that: “as long as will can still unplug them [AI/robots], I don’t think we have anything to worry about … we exacerbate things simply because they’re exciting.” 

Another advocate of the technology is UK Technology Secretary, Peter Kyle, who, according to a report in the Guardian, has urged employees to “act now” and embrace technology, or delay and risk being “left behind.” 

Lobbying for government industry initiatives to train 7.5 million workers in AI by 2030, Kyle has stated that just 2.5 hours of training can bridge generational gaps in AI adoption. 

However, “act now” or risk being “left behind” is a reactive statement, and it brings us to a central tension of our age: innovation cuts both ways. 

The duality of AI 

As AI increases output and insight, with agentic AI congratulating itself on doing a “good job” it is slowly diminishing the need for human input. From a business perspective, they will undoubtedly benefit from sharper efficiencies and having an edge in competition – but at what cost to the workforce? 

We’ve already read about concerns of how AI will potentially slash entry-level positions with up to 20% of the workforce in the US becoming unemployed. And if history tells us anything, it’s that each industry advancement – from the loom to the television to the Internet – carried with it the weight of disruption to workforces alongside its progress. However, AI feels a little different. It has the potential to further exacerbate bias – algorithms having been trained on inherently skewed data, plus its hallucinations could spread misinformation. In fact, Google’s new AI Mode feature in the US synthesises information from various sources to provide a cohesive answer, meaning there are no backlinks to check the validity of the information. 

For me, AI is interesting and I believe that many companies look at it through a lens that considers a future with only one possible output – the positive side of AI; such as faster working processes, faster insights, and business cost reduction. However, the negative sides that counterbalance this are: a very real potential for job losses, no emotional intelligence, bias, and security concerns to name a few. 

Amazon has been vocal on how it retrains its staff into engineering positions, enabling them to grow and evolve with a company. However, there are two concerns here. One, if AI makes practices faster and can save a company money whilst providing insights – then it stands to reason that a big share of that workload will be given to AI and companies won’t need an entire workforce – presumably on a higher salary to reflect the increased skills – to do the work, no matter how much more skilled they are and so, job losses will be a factor. And two, and maybe a simpler question. What if the employee enjoys their job and simply doesn’t want to retrain. 

From my own current perspective, I feel that a Pandora’s box is now open and the lid can’t be closed. Generative AI has catapulted us forward into very new territory, and companies are racing to deploy it faster than ever – because technology and consumer demands are evolving faster than ever. There are many arguments for and against AI, and now with agentic AI also on the horizon, the question I find myself asking is – just because we can, should we?