Goldman Sachs wrote an article, and they have 7 AI predictions for 2026. In today’s episode, I want to share them with you and explain how they might apply to us.
Goldman Sachs CIO, Marco Argenti, has 40 years of experience in technology. He says 2025 was the biggest year of change he has seen in his career. And he believes 2026 will be bigger still. He has made 7 predictions for how AI evolves this year.
- AI models become operating systems (not just chatbots)
- Memory becomes the key engineering priority, not model size
- Personal AI agents arrive for consumers
- Mega alliances reshape the industry (winner-takes-most dynamics)
- The gigawatt ceiling: energy becomes the binding constraint
- A tale of two nations: US vs China AI race intensifies
- Workers who adapt will thrive; curiosity beats expertise alone
Trend 1: AI models stop being chatbots and become operating systems
Until now, AI has mostly been something you ask questions of. That is changing. In 2026, these models start behaving more like a computer’s operating system. They access tools, browse the web, open files, and get things done. You stop typing prompts and start giving instructions. The companies that own the best models will effectively own the new platforms everyone else builds on. That is a big shift in where power sits.
Trend 2: Memory becomes more important than model size
The last two years have been a race to build bigger, smarter models. Argenti says that race is changing direction. The models know an enormous amount. What they are still poor at is remembering you specifically, your context, your previous conversations, and your preferences. The focus in 2026 shifts to fixing that. Less about raw intelligence, more about relevance to you.
Trend 3: Personal AI agents arrive for consumers
Argenti’s analogy here is a good one. Using AI right now is like getting turn-by-turn directions. You prompt, it responds, you act. What is coming is more like a sat-nav. You say where you want to go, and it handles the route. The things you currently do across a dozen different apps manually, booking, searching, comparing, and scheduling, start getting handled for you. We have already seen early versions of this. It goes mainstream in 2026.
Trend 4: Mega alliances reshape who can compete
Scale matters enormously in this space, and the way companies are securing scale is through major partnerships. Argenti’s prediction is that the alliances forming in 2026 will be so large and so self-reinforcing that only a handful of players will realistically be able to compete at the frontier. He compares it to aerospace, where the barriers to entry are so high that the industry ends up as a duopoly. We are already watching it happen. Disney and OpenAI. Goldman and Anthropic. Meta is acquiring Manus. For organisations buying AI, the pool of truly independent vendors is getting smaller.
Trend 5: Energy becomes the binding constraint
This is the trend that gets the least attention and deserves the most. Goldman Sachs Research projects that data centre power consumption will nearly triple by 2030 compared to 2023. The physical infrastructure required to run AI at this scale takes years to build. Argenti calls it the gigawatt ceiling. In practical terms, access to power is becoming as strategically important as access to capital. For any organisation with net-zero commitments, the energy cost of the AI you are running is no longer a question for your IT team. It belongs in the boardroom.
Trend 6: A two-horse race
The geopolitical picture is sharpening. Argenti puts it plainly: the US and China both have a credible shot at leading on AI, and the gap between them is narrowing. For Europe, this is precisely why the sovereign AI conversation matters. France is embedding Mistral in its military infrastructure. The EU AI Act. These are not just policy moves. They are attempts to ensure Europe is not left importing its AI from elsewhere and operating under someone else’s rules.
Trend 7: Curiosity beats expertise
This is Argenti’s most direct prediction about people. The workers who do well are not necessarily the deepest experts. They are the ones willing to rethink how they work. He says an expert who clings to old habits will be outperformed by someone less experienced but genuinely curious about working differently. Coming from the CIO of Goldman Sachs, that is worth sitting with.
Know it, use it, govern it. That is my framework, and it maps directly onto what Argenti is describing. The knowing is about understanding what these tools can actually do. The using is about building them into your work deliberately. The governing is about the harder questions: energy, sovereignty, vendor dependency, and the people in your organisation who need to keep pace.
Seven trends, one direction. In 2026, AI is no longer something you experiment with. It is the environment you are working in.
If you want to build your AI literacy and skills, I can help.
📚 Grab AI Playbook on Amazon, the perfect podcast companion.
📰 Subscribe to Human in the Loop, my LinkedIn Newsletter
🎓 Take the CPD Accredited AI Playbook Diploma
📞 Book a call at digitaltraining.ie
About the Podcast Host
AI and Communications Specialist | Author | Podcaster | International Speaker
Joanne Sweeney is a recognised leader in digital marketing and digital transformation, specialising in the practical application of AI to enhance communication strategies. The founder of Digital Training Institute, Joanne, empowers leaders and communications professionals to leverage AI responsibly and effectively, ensuring transparency, trust, and tangible results.
She completed the Harvard Kennedy School Executive Programme in AI and Public Policy this summer before launching her fourth book AI Playbook, in September. Joanne has also published two editions of Public Sector Marketing Pro, which were preceded by her first title, Social Media Under Investigation Law Enforcement and the Social Web.
Her books have become go-to resources for governments, public sector and policing agencies, as well as marketing and communications professionals, all navigating the challenges of the changing Internet.
Joanne is the host of the AI SIX Podcast, a daily six-minute podcast that provides actionable insights, expert interviews, and practical advice to help professionals navigate the new world of work.
