For the last couple of years we’ve tried to predict what’s coming next in AI. It’s a bit of a fool’s game given how fast this industry moves… But we’re on a roll, so we’re doing it again. In this edition of What’s Next in Tech, discover what’s next for AI in 2025.
What’s coming next in the fast-paced world of AI? Join MIT Technology Review’s editors on January 16 for 5 AI Predictions for 2025, a special LinkedIn Live event exploring transformative trends and insights shaping the next twelve months of AI and business. Register for free today.
- Generative virtual playgrounds: If 2023 was the year of generative images and 2024 was the year of generative video—what comes next? If you guessed generative virtual worlds (a.k.a. video games), high fives all round. We got a tiny glimpse of this technology in February, when Google DeepMind revealed a generative model called Genie that could take a still image and turn it into a side-scrolling 2D platform game that players could interact with. In December, the firm revealed Genie 2, a model that can spin a starter image into an entire virtual world. Other companies are building similar tech.
- Large language models that “reason”: The buzz was justified. When OpenAI revealed o1 in September, it introduced a new paradigm in how large language models work. Two months later, the firm pushed that paradigm forward in almost every way with o3—a model that just might reshape this technology for good. Most models, including OpenAI’s flagship GPT-4, spit out the first response they come up with. Sometimes it’s correct; sometimes it’s not. But the firm's new models are trained to work through their answers step by step, breaking down tricky problems into a series of simpler ones. When one approach isn’t working, they try another. This technique, known as “reasoning” (yes—we know exactly how loaded that term is), can make this technology more accurate, especially for math, physics, and logic problems. It’s also crucial for agents.
- It’s boom time for AI in science: One of the most exciting uses for AI is speeding up discovery in the natural sciences. Perhaps the greatest vindication of AI’s potential on this front came last October, when the Royal Swedish Academy of Sciences awarded the Nobel Prize for chemistry to Demis Hassabis and John M. Jumper from Google DeepMind for building the AlphaFold tool, which can solve protein folding, and to David Baker for building tools to help design new proteins. Expect this trend to continue next year, and to see more data sets and models that are aimed specifically at scientific discovery. Proteins were the perfect target for AI, because the field had excellent existing data sets that AI models could be trained on. The hunt is on to find the next big thing.
Read the full story for more on these three predictions, as well as two additional things our team anticipates will happen this year in the world of AI.
- Anthropic’s chief scientist on 4 ways agents will be even better in 2025 — The hottest topic in AI is only just getting started.
- AI means the end of internet search as we’ve known it — Despite fewer clicks, copyright fights, and sometimes iffy answers, AI could unlock new ways to summon all the world’s knowledge.
- The biggest AI flops of 2024 — From chatbots dishing out illegal advice to dodgy AI-generated search results, take a look back over last year’s top AI failures.
Image: Stephanie Arnett/MIT Technology Review | Lummi
Professor in Innovation Management | Global Futurist | Author of 30 books on Purpose-Driven Innovation, AI, Governance, Design, Leadership, and Sustainability | Endorsed by Donald Trump: "TO HUBERT, ALWAYS THINK BIG!"
1 个月This is next in AI https://hkrampersad.wordpress.com/2025/01/11/purpose-driven-ai/
OK Bo?tjan Dolin?ek
Director at Omnikare Ltd
1 个月The search for the ‘next big thing’ idea can be given a focus. The idea rests on a confluence of circumstances, competencies and content that can be brought to play to create an ecospace for universal participation of enterprises globally. The business space is open for an ecosystem that enterprises inhabit as a matter of course, a type of ‘mycorrhizal’ network. A new door to the web that offers a dynamic and transient space for enterprises to inhabit.? Such an ecospace needs a taxonomy and ontology, a domain specific foundation that unambiguously offers recognition of enterprises for what they are – consumers and producers of a spectrum of services and products that emanate from ‘value-added’ activities. A vocabulary is needed, one that could only evolve from a Big Tech initiative that defines the mission – ‘engage with over half a billion enterprise that populate the global economy like fireflies’. The strategy could then manifest as a system design that would feed-off a live, evolutionary, bottom-up ‘vocabulary’ system that employs hybrid competencies – generalist and specialist, human and digital, laying down in the process the highways and country roads of a transitory landscape. ?A home for enterprises. All of them.
The AI landscape continues to evolve at an astonishing pace, and the insights shared in MIT Technology Review’s “What’s Next for AI in 2025” are both thought-provoking and inspiring. ?? From generative virtual worlds to reasoning-enabled language models and groundbreaking contributions in scientific discovery, it’s clear that AI is set to redefine the boundaries of possibility in the coming year. These trends highlight not only the innovation happening within AI but also its profound potential to solve real-world challenges.