Video summary

Ответы начинающим программистам про ИИ и эмиграцию

Main summary

Key takeaways

Educational

Main ideas / lessons from the video

The speaker frames the video as five questions aimed at beginner programmers or people trying to enter programming.

1) How to use AI in professional programming/engineering work

  • The speaker uses AI only for routine automation, not for core technical development.
  • Example use case:
    • Automating tedious code such as writing a Windows API function that requires many arguments (e.g., querying/reading data from the registry).
    • AI can draft an initial “framework,” after which the speaker adjusts and fixes it manually.
  • Why AI isn’t useful for the speaker’s main work:
    • AI needs a strong knowledge base to produce correct results.
    • In the speaker’s domain (scientific/research-oriented engineering), the AI’s knowledge is described as weak.
    • Generated output is described as “crap,” and correcting it can take twice as long as writing the code oneself.
  • Time/impact estimate:
    • AI is used for about ~5% of tasks/time.
  • Forecast:
    • Over the next 5 years, the speaker predicts AI will not be used broadly in their field as a full tool—especially not for processing sound/video/images in their context.

2) Which programming language to learn for robotics

  • The speaker criticizes the idea that there’s a single “correct” programming language for robotics.
  • Key point: Robotics is not just programming—programming is only a minority of the overall robotics effort.
  • Practical labor split (as stated):
    • Building the robot’s mechanics/kinematics/dynamics, manipulators, chassis design, etc. dominates the work.
    • Programming the robotics “brains” is estimated at ~5–10% of the labor intensity.
  • Implications:
    • If you like robotics/technology, you might not end up mainly doing programming.
    • If you join a team building the robot (including everything around it), any language can be appropriate depending on needs.
  • Language options (low-level to high-level):
    • Assembler
    • C, C++
    • Python
  • Where/how code runs:
    • Robots can range from primitive microcontrollers to single-board computers like Raspberry Pi.
    • On Raspberry Pi-like systems, you can run a (modified) Linux-style OS and use many languages.
    • Platforms mentioned: Arduino (modified C) and Raspberry Pi.
  • Industrial note:
    • For “industrial application,” the speaker implies you often end up with assembler for cheap products, with C/C++ commonly used elsewhere.
  • Final advice:
    • Clarify your goal: robotics may not be primarily programming.
    • Choose tools/languages based on competence, budget, and customer requirements.
    • If working personally (no client), choose what you already know best.

3) Prospects for AI development in Russia (and how it “should” develop)

  • Main claim:
    • There are no prospects for broad AI development in the Russian Federation.
  • Reason (as framed by the speaker):
    • The ruling elite allegedly doesn’t need AI and isn’t interested (presented in a strongly negative, political tone).
  • Acknowledged counterpoint:
    • The speaker says they know people working on AI in Russia (including Sber and work in banks).
    • The issue is scale compared to other countries.
  • Comparison metaphor:
    • AI progress in America/China vs Russia is compared to “an elephant and a pug” / “a little dog can at least bite an elephant”—meaning Russia’s efforts are far behind.
  • What the speaker would do instead:
    • Develop AI tightly connected to classical industries, especially:
      • Mechanical engineering
      • Construction
      • Agriculture
      • The “real sector” of the economy
    • Use AI for practical tasks such as:
      • Decision-making in industrial contexts
      • Machine vision
      • AI/automation in production and engineering processes
  • What they reject:
    • Avoid focusing on AI hype products like chatbots or superficial flashy robotics.
    • They also criticize “anthropomorphic robots for hype” (e.g., viral China examples).
  • Market argument:
    • The speaker claims there’s no real market in Russia for those alternative AI products and that they’re not necessary for national development.

4) Is there salary differentiation by programming language?

  • The speaker argues that salary “rankings” by language are misleading.
  • Why rankings are unhelpful (as stated):
    • They can be used to manipulate job seekers and “mess with their heads.”
    • Rankings rely heavily on vacancy data, which the speaker claims is often fake or inflated.
    • Companies may game rankings by posting many vacancies in a language (e.g., Perl, Fortran) with high salaries.
  • What determines pay instead (speaker’s “rule”):
    • Pay depends less on the language itself and more on:
      • Whether employers have orders and money
      • Whether the client’s industry is healthy enough to fund software/services
  • Economic condition point:
    • Even if programmers were paid well earlier, client industries can collapse, shrinking budgets.
  • Advice:
    • Don’t chase language rankings.
    • Identify industries that are doing well and have money.
    • Determine what tools/solutions they need.
    • Choose technologies that solve those needs.

5) “Is IT hiring completely dead?” (VRF hiring question; interpreted as IT hiring)

  • The speaker says IT hiring is not dead.
  • Argument against “hiring is dead” narratives:
    • YouTube/bloggers discuss trends regardless of truth; their content is driven by what gets views.
    • Many bloggers monetize content (ads, monetization platforms), so they may hype whatever is popular.
  • Core claim:
    • The real issue is a broader labor market problem:
      • global recession / economic decline
    • Hiring problems affect more than IT.
  • Industry resilience:
    • IT workers will likely “survive” downturns.
    • Other industries may suffer more (speaker references resources like coal/forestry as essentially bankrupt, plus retail/agriculture as problematic).
  • Social media critique:
    • It’s fashionable to complain that IT is bad.
    • Bloggers may ignore non-IT realities because it doesn’t fit their content ecosystem.

6) Immigration and which country to choose (Canada discussion within the immigration theme)

  • The speaker discusses emigrating mainly toward Canada, then argues against it.
  • Canada as described:
    • Not like it was ~15–20 years ago (a “then vs now” contrast).
    • Issues mentioned:
      • Lower salary levels than expectations
      • Real estate stagnation with high prices
      • Broad social problems (e.g., homelessness, drug addiction)
  • Cost barrier and practical constraints:
    • To move, you need either:
      • significant wealth upfront (e.g., buying a home without a mortgage), or
      • be a rare super-specialist in demand by a specific Canadian employer
    • The speaker argues most people can’t meet these requirements.
  • Why early emigrants often stay:
    • “Sunk costs” (owning real estate); selling doesn’t produce enough to relocate again (e.g., to the USA).
  • Final stance:
    • Immigration is framed as realistic mainly for the “Anglo-Saxon world” (England/USA/Australia mentioned).
    • The speaker suggests it’s suitable only for people with special circumstances and “strong spirit,” not for most.

Speakers / sources featured

  • Alexander Grigory (sole speaker; identified in subtitles as “Alexander Grigory is with you again.”)
  • Organizations mentioned (not as speakers):
    • Sber
    • Russian banks
  • Platforms mentioned (not as speakers):
    • YouTube (and other platforms briefly implied)
    • Telegram, Rutube (as monetization/content platforms)
  • Countries/regions mentioned:
    • Russia, USA, China, Canada, England (UK), Australia

Original video