April 2024 Recap: AI-Augmented Collective Intelligence
Gianni Giacomelli
Researcher | Consulting Advisor | Keynote | Chief Innovation / Learning Officer. AI to Transform People's Work and Products/Services through Skills, Knowledge, Collaboration Systems. AI Augmented Collective Intelligence.
The last twelve months have been full of insights and surprises in the ACI space. Here's a curation of the most interesting ones, and a reflection on the trends.
"People vs machines" continues to be a big (and somewhat misleading) narrative - and evidence of complementarity is emerging. Some organizations struggle with change, but the target function, and the tools to get there, are becoming clearer. Workers also struggle with that change, and the guidance and infrastructure we give them to support them are not yet fit for purpose. Yet, there are reasons for huge optimism too, when it comes to harnessing the knowledge and intelligence around us - including for progress, and science. Much of the future's outcomes depend on our ability to design the right human-machine synergies - that continues to be the job at hand, and the reason why this collection of perspectives can be of help.
1?? The big picture
2?? AI-supported organizational transformation to make enterprises truly collectively intelligent
3?? Skills in the age of AI
4?? Knowledge in the age of AI
5?? Innovation in the age of AI
6?? AI and its engagement with humans
7?? AI using collective intelligence architectures
1?? The big picture
?? Dozens of scenarios where AI-augmented collective intelligence could make a significant positive impact on our world
?? Experts have been revisiting the timelines that AI will take to beat certain human skills. 10 to 30 years earlier
?? AI "vs" or "+" humans: Turing Trap 's concept, essential knowledge now. But "one person's automation is another's augmentation ".
?? Technological change and top Labor Incomes, a study
?? A useful interview on Generative AI between Azeem Azhar and Ethan Mollick (video)
?? GenAI is threatening some online communities
?? Are we collectively shifting away from a culture of progress?
?? The pace of innovation may be slowing when people age
?? Executives' view on risks . Lots to be addressed through ACI
?? The climate transition requires more than just technical skills, and professionals across a broader spectrum must interoperate better
?? Collective intelligence tracking the exploitation of natural resources
??? Democratic decision-making is a process - one that orchestrates collective intelligence (video)
2?? AI-supported organizational transformation to make enterprises truly collectively intelligent
?? A now-classic study on the impact of generative AI on consultants' work, with implications well beyond it. And here
???? GenAI models are still too inaccurate and unreliable for many processes. But so are humans, in isolation. Practices and methods to solve this fundamental issue
? Generative AI requires organizational transformation to take hold
?? Distance from the workplace has increased - much of that change is not because of people moving farther in the same area, but rather due to people moving very far away, and using technology to work
?? Using AI to understand where AI can insert itself in processes
?? A nice example of the structures of collective intelligence - and with implications beyond individual countries
?? Leveraging AI to improve business collaboration
??? How much diversity do you need in human networks?
?? The next generation of videoconferencing. Video . Also, check out Marc Zuckerberg's interview with Lex Friedman using Meta's experimental VR gear.
? 120 years ago. One big lesson. Can we give AI its Kitty Hawk moment ? Some management lessons
?? What the history of Microsoft Excel can tell us about AI taking our jobs
?? Pay gender gaps, and their implication for the future work knowledge management, and AI
?? The organizational challenges that slow down the progress of enterprise generative AI
? Can organizational design truly embrace AI? Podcast
?? Physical work structures have an impact on the collective intelligence of organizations. There's still a lot of debate, emotions, and confusion about it. Here is a perspective , and another
?? Companies become collectively more stupid when they force their people to work synchronously more often than they need to
3?? Skills in the age of AI
?? How many jobs will AI really take?
???? A large challenge with AI is that we aren’t guiding our workforce to what they should expect. Video and essay
?? A case study on building a cost-effective, agile learning infrastructure that helps people prepare for an AI world. And an HBR article written by Ryan Roslansky on the same topic.
?? Students becoming dependent on AI learning tools, which highlights risks we need to design against
领英推荐
?? A reminder that the T-shape concept (or similar) is crucial in thinking about skills for AI's age
?? Critical thinking (in all its forms) is an essential skill for humans as we will work alongside AI in the future. This is another study that shows that we are not ready
? A great list of what AI can do for human skills
?? Early examples of using AI to engage adult learners
?? What can Leonardo da Vinci teach us about learning in the age of AI?
4?? Knowledge in the age of AI
?? An example of the future of AI + Knowledge Graphs
?? Using AI to improve our scientific and technological research
?? The challenges are interdisciplinary , and so are the solutions. Many innovators in these spaces look at solutions in a siloed way. Apart from the mindset change, support from GenAI could help find systemic root causes, and recombine solutions from various spaces.
?? How to rebuild organizations' knowledge infrastructure with today's technology. A segmentation
?? Good review of hallucinations in LLMs
???? A (relative) long time ago, one of the first general discussions about the usage of knowledge graphs for generative artificial intelligence
5?? Innovation in the age of AI
?? Generative AI can ideate harder if we design it to do so. Also here
?? The collective intelligence of AI models seems to overcome individual model limitations on business strategy
?? Do LLMs generate less diverse ideas compared to humans? And, an early study that investigated both the average quality as well as the variance of human and AI creativity
????Should we wait a little longer until the AI tools are better, given the pace of change?
?? There's a clear opportunity to explore the synergy between humans and AI in the ideation, and more generally the innovation process. Looking at psychedelics as an analogy might yield new, practical perspectives
???? Compared to humans, GenAI still struggles with abstract concepts that it has not been trained on. This may change with time
?? AI-powered Superminds for innovation (video )
???? Another reminder of the utility of GenAI for innovation. The upshot: this is not a substitute for humans (in this case, peer reviewers) but it can be a great complement
?? An interesting paper showing how LLMs can help combine ideas across different fields
? We can get machines to help us ideate better. The MIT Supermind Ideator is one of the first examples of a new innovation stack, where an intentional process leads to a user experience that leverages AI well. Paper
?? After a "superstar" scientist dies, their field accepts contributions from entrants from other disciplines more freely. Technically, strong networks that share similar frames of reference inhibit collective intelligence across weaker ties. Paper
6?? AI and its engagement with humans
?? An interesting concept: AI that antagonizes you (within guard rails) to make you sharper - and improve the collective output
?? LLMs display human behavior traits, which helps them interoperate with humans
?? Will AI make us smarter or duller ?
?? How to amplify our cognition with ACI (podcast)
?? Study shows that pairing humans with GenAI is already able to produce faster, better output
??? We want LLMs that ask questions to refine their assessment not only give answers to instantly gratify us. And we want them to show empathy. Here’s a study that explores that
?? A study on the use of robots in helping baseball umpires shows that the combined human-machine duo improves performance over humans alone, especially for lower-skilled humans
?? Study highlights the complementarity between LLMs and humans in peer-reviewing scientific papers
?? Sycophancy (uncritical agreement with others) is a problem for collective intelligence. So research studying it in LLMs is important. Here
?? On the risk of humans falling asleep at the wheel with AI
7?? AI using collective intelligence architectures
?? The design space of "multiple GenAI , combined", comprising varied architectures like Mixture of Experts, and likely exemplified in GPT-4, is being increasingly explored
??? Multiple AI agents, when combined in specific organizational architectures and workflows, may give rise to improved performance
?? What part of the world's collective intelligence has fueled the knowledge of language models?
That's all for now, see you in a few months.
Feel free to share this newsletter and encourage others to subscribe. And for more timely updates, follow my feed .
...
7 个月What a comprehensive look into the world of AI. Excited to dive into these articles. ??
AI Experts - Join our Network of AI Speakers, Consultants and AI Solution Providers. Message me for info.
7 个月Sounds like a comprehensive overview of AI's impact on various aspects. Excited to delve into the articles.
Assistant Vice President - ERM & AI Enthusiastic
7 个月It was indeed a good read. Though I feel we are still at distant on deploying actual AI. ACI is the focused that an organization should adopt and work towards it. LLMs are also being created offline to maintain data privacy any thought on that? Whether do you see that is going to be new BAU?
Empowering L&D with AI
7 个月state of the art! what a rich resource this is Gianni Giacomelli, thanks for organizing this for the rest of us.
Founder and CEO: Analytics-Based Performance Management LLC; Expert in ABC, EPM/CPM, Profit Analysis, Budget, Analytics
7 个月Gianni ... Thanks. A well written and relevant article.