I had the honor of giving a keynote at the International Conference on Machine Learning in Seoul last week titled “What will be left for us to work on?” I addressed the widespread anxiety about how we should adapt as AI capabilities increase. I was thrilled by the talk’s reception, so I have made my slides available, annotated with a lightly edited transcript: https://t.co/vNgRJCL57B
I made three arguments. First, the "AI as Normal Technology" framework is a correct and useful as a way to think about AI’s impacts, unless and until there is some future discontinuity such as through recursive self-improvement. Second, even though we should take recursive self-improvement seriously, there is no milestone that companies might achieve in the lab that will suddenly put us all out of work. Third and finally, jobs of the future will be radically different, and a lot of adaptation will be needed. I shared my thinking about what this might look like and ended with a vision of human/AI “co-superintelligence”.
Thank you for dropping by and initiating this interesting conversation. I am pretty much sure that even with a full AGI system we will still find our own space and purpose as defeating humans can be a long way to go for machines. Hope you enjoy the aura of AI makers next few days
My very first conversation at #acl2026 is with @mohitvaishnav who gave me the best reason why computer science is still needed even in a world where AI can code:
“We built the machine and now we need to learn to master it.” Says people who have computer science degrees have a better knowledge base upon which to do that.
After he said that I asked him if I could record and this is what we recorded.
I have already had some amazing conversations.
More talk here about spatial computing than I was expecting.
One attendee told me why: the people who make Large Language Models e that one path to improving them is to give them eyes, and other sensors, and let them move around the world on their own.
My brain is tired, because everyone I meet knows something I don’t about how AI works.
Then we all went out to watch San Diego’s fireworks show, billed as the biggest on the West Coast.
Happy 250th America. I am blessed to lower the IQ of every room I walk into the next few days.
Does smarts rub off? Only if you really listen. :-)
Excited to share that our paper on VLM perception-reasoning bottlenecks will be an ORAL presentation at #ACL2026! 🌴
If you are in ACL and want to chat about visual reasoning, let's connect! ☕️
https://t.co/atWZlZfhaA
#ACL2026#VLMs#VisualReasoning#NeuroSymbolicAI#NLP#AI
#US#AI policies are a "#governance mirage." 🌫️ Without technical logs for #EUAIAct compliance, firms risk fines up to 7% of global turnover. Policy isn't proof: you need #infrastructure.
#Casestudy: https://t.co/uqpz82O5G1
Website: https://t.co/yFF81iGEtm
#ISO27001#GDPR#GRC
January 2026 exposed major enterprise risks as #Warlock exploited a "forgotten" #VM at SmarterTools, chaining #CVE202623760 with LotL tactics to bypass #MFA and steal 1.2 million records.
Casestudy: https://t.co/EMbOkiF1JF
Website: https://t.co/n73kmdjzAe
#Storm2603#Ransomware