Artificial Intelligence

Top Wonderings Q/A

  • Human vs Machine: What is the goal of AI? Why do we need a machine to have intelligence?

    • Use machine as a tool, to assist people achieve goal.
    • Will machines have goals, and some day they use human as tools to achieve machine’s goal?

  • Human vs AI: What do we want to do using AI?

    • Some really difficult tasks for human:
      • Compute $\pi$
      • Go to the moon –> space craft;
      • Limited memory and processing power –> Big Data storage and analysis;
    • to release human from labor:
      • driving –> autonomous driving car, smart road?;
      • manufacturing –> hand-arm robitic pipelines;
      • farming –>

  • Artifitial Intellegence vs. Human intellegence: Now we know many human intellegence (object recognition for example) does not yield 100% correct ratio, it seems not good to follow human’s principle to do the same task and we should find new ways that can achieve a much better result, yes?

    • e.g., counting animals using fingers and toes is hard sometimes, so we find new tools (such as stones?) to count instead of using our own organ.
    • e.g., safeguards (or smart-doors) of a building cannot recognize everyone’s face in order to determine who is allowed to be in, so we have ID card for him to check instead of checking our face’s image.
    • e.g., if a computer cannot recognize a car/wheel/lanes/stone/blocking on the road, why not just create new feature for each of them so that computers can detect them much easier with higher correct ratio?

  • How much Image and Voice processing based AI can be useful?

    • Image/Voice as information media: does this solution as communication media have a ceiling limitation from its essence? (Human handle image/voice better than machine, but can still have mistakes)

      • ceiling limitation of correct ratio, communication efficiency, etc.
      • ceiling limitation of ??

    • What can be the enhancement of intelligence/learning if image/voice themselves is not enough?

      • The language. We listen, hear, watch, but also read, write language as special encodings of information. Language is a media that can embed information for better communication. Draw a picture on the stones is an intermediate media form. But language has been developed to be a special media that can describe much more information than pictures or voices can represent. So, to enhace the machine that does not have good image/voice processing power, a powerful language can be used, especially for communication between machines (language does not work well cross species, but voice and image do.)
      • Physical touch.
      • Chemical reactions. Smell/Taste.
      • Crafted tools: microscopes that can help us find living cells;

    • Any other usage of Image/Voice rather than as a information media of communication?

  • How to defend against (or recognize) Faked digital information by AI?

    • Faked/Synthesized photos, voices, videos, etc.
    • Digital signature? NO?

References:

  • A list of PDF Books about Machine Learning on github

Tasty critiques on 5G and AI

More

  • Autonomous Driving Cars

    • References: V2X: Vehicle to everything More

  • ML

    • References: reference More Reading Notes on Dr. Mi Zhang's Publications References: Mi Zhang - Publications Publications in 2020 Mutual Net ECCV’20: MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution Distream SenSys’20: Distream: Scaling Live Video Analytics with Workload-Adaptive Distributed Edge Intelligence WiFi SenSys’20: WiFi See It All: Generative Adversarial Network-augmented Versatile WiFi Imaging. SecWIR MobiSys’20: SecWIR: Securing Smart Home IoT Communications via WiFi Routers with Embedded Intelligence.

  • History

    • References: 人工智能简史（图灵出品），尼克，人民邮电出版社，2017-11 人工智能是一门新学科，历史的读物并不多。波登的《认知科学历史》（Boden 2

Created Nov 23, 2019 // Last Updated Aug 31, 2021

If you could revise
the fundmental principles of
computer system design
to improve security...

... what would you change?