China’s Manus AI is showing an excellent demo of agent capabilities. You can toggle between four applications (Search, browser, coterminal and document editor).
David Sacks says AI agents will be connect to all of hour SAAS (Software As A Service) applications. Agents will interact with dozens of apps. There is a new Agent standard called MCP where people will be able to connect to all applications. This will be a very standard UI paradigm
AI Agents will be able to do everything on a computer that you can do.
It is running behind the scenes. There is new work and not just a replacement. 90% of the work is work that would not have been done. Verifications and other work that would not have been done.
How do you make reliable enterprise AI solutions ?
There are three exponential improvement axis. (Per David Sacks).
The algorithms for the models. Not just faster but qualitatively better.
The chips are getting better.
The data centers (the numbers of chips and how they are connected together).
Each dimension could be 100X better in 4 years for a combined 1 million times better training and AI compute capability in 4 years.
I think the XAI data center will be at 1 million chips at the end of this year. This will be a ten times increase from 100,000 chips just 6 months ago.
The large AI data centers and chips could unlock more chip and algorithm improvement. An extra 100X to 1000X unlock from acceleration could enable 100 million to 1 billion times increase in AI compute capability in 4 years.
Coding and math where there is easy validation and proofs are making great advances.
Get to Enterprise Quality and Low Error Rates
There is some 90% first pass correctness but 10% error rates of reviewing a large document load to get searched 40 data fields.
There now needs to be work to reduce the error rate to enterprise acceptable levels.
The best people need to be in Quality assurance. There needs to be running multiple passes and working out the verifications and sometimes a need to break up the data load into smaller pieces.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
I am still hoping that avoiding the use of AI for the online apps will remain.
I’m sure the Chinese will be thrilled:
“As per a leading aerospace engineer quoted in the report, China is using its homegrown AI assistant in making the most advanced warplanes.
Wang Yongqing, the lead designer at the Shenyang Aircraft Design Institute, reportedly told state-owned Chinanews.com that his team was already using DeepSeek to make new technology for the planes.”
See:
https://www.livemint.com/technology/china-using-deepseek-ai-to-design-of-advanced-warplanes-lead-engineer-says-providing-new-ideas-and-11746367653548.html