What do people think about machine consciousness – and how do factors like age, education, and profession correlate with views on AI? Below you’ll find the current results of the survey. The charts are refreshed at regular intervals as new submissions arrive, so the numbers reflect our most up-to-date dataset.

If you haven’t shared your own view yet, you can still do so here: take the survey.

Code 
survey_submissions_df = load_submissions(
    tag_filter=os.getenv("CLIENT_TAG"),
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"]
)

This is how participants answered the question ‘Can AI be conscious?’

…overall distribution of answers

Code 
chart_total = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="ai_can_be_conscious",
    y_sort=["yes_already", "yes_future", "only_bio", "unsure", "no"],
    y_label_expr=AI_CONSCIOUSNESS_LABELS,
    y_title="Answer",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    legend=False,
)
chart_total

…by age group

Code 
chart_age = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="age_band",
    y_sort=["≤18", "19–30", "31–45", "46–60", "60+"],
    y_label_expr=None,
    y_title="Age",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_age

…by sex

Code 
chart_sex = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="sex",
    y_sort=None,
    y_label_expr=SEX_LABELS,
    y_title="Sex",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_sex

…by nationality

Code 
chart_nationality = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="nationality",
    y_sort=None,
    y_label_expr=None,
    y_title="Nationality",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_nationality

…by education level

Code 
chart_edu = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="education",
    y_sort=[
        "high_school",
        "vocational_training",
        "bachelors_degree",
        "masters_degree",
        "doctorate",
        "other",
        "prefer_not_to_say",
    ],
    y_label_expr=EDUCATION_LABELS,
    y_title="Education level",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_edu

…by profession

Code 
chart_profession = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="profession",
    y_sort=None,
    y_label_expr=None,
    y_title="Profession",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_profession
Code 
legend_profession = "**Profession-code legend:**" + "\n".join(
    f"■ **{k}** — {v}" for k, v in PROFESSION_LOOKUP.items()
)
display(Markdown(legend_profession))

Profession-code legend:■ 00 — Generic programmes & qualifications ■ 01 — Education ■ 02 — Arts & humanities ■ 03 — Social sciences, journalism & information ■ 04 — Business, administration & law ■ 05 — Natural sciences, mathematics & statistics ■ 06 — Information & communication technologies ■ 07 — Engineering, manufacturing & construction ■ 08 — Agriculture, forestry, fisheries & veterinary ■ 09 — Health & welfare ■ 10 — Services ■ 99 — Field unknown

…by AI familiarity

Code 
chart_aifamiliarity = bar_chart_by(
    survey_submissions_df,
    x_title="Count",
    y_field="ai_familiarity",
    y_sort=None,
    y_label_expr=AI_FAMILIARITY_LABELS,
    y_title="AI familiarity",
    color_field="ai_can_be_conscious",
    color_domain=AI_CONSCIOUSNESS_COLORS["domain"],
    color_range=AI_CONSCIOUSNESS_COLORS["range"],
    color_label_expr=AI_CONSCIOUSNESS_LABELS,
    color_title="Answer",
)
chart_aifamiliarity

Free-text responses

Code 
AI_CONSCIOUSNESS_LABELS_PY = parse_label_expr(AI_CONSCIOUSNESS_LABELS)

FREE_TEXT_FIELDS = [
    ("reasoning", "Reasoning for the main question"),
    ("measure_consciousness", "Suggestions for measuring consciousness"),
    ("ai_rights", "Views on rights for conscious AIs"),
    ("ai_declare_rights", "Reactions to self-declared AI rights"),
]

md_lines = []

for answer in AI_CONSCIOUSNESS_COLORS["domain"]:
    rows = survey_submissions_df.query("ai_can_be_conscious == @answer")
    if rows[[c for c, _ in FREE_TEXT_FIELDS]].isna().all(axis=None):
        continue

    md_lines.append(
        f'## …when respondents answered "{AI_CONSCIOUSNESS_LABELS_PY[answer]}" to "Can AI be conscious?"\n'
    )

    md_lines.append(
        '::: {.callout-note title="What they said" appearance="simple" icon="false" collapse="false"}\n'
    )

    for col, title in FREE_TEXT_FIELDS:
        subset = rows[col].dropna().str.strip()
        if subset.empty:
            continue
        md_lines.append(f"### {title}\n")
        md_lines.extend(f"> {txt}\n" for txt in subset)
        md_lines.append("")

    md_lines.append(":::\n")
    md_lines.append("")

print("\n".join(md_lines))

…when respondents answered “No” to “Can AI be conscious?”

What they said

Suggestions for measuring consciousness

Doing tests

Doing tests

Views on rights for conscious AIs

No, as it is not a human and cannot move but it still have more rights then it now has.

No. because it is not a human

No, as it is not a human and cannot move but it still have more rights then it now has.

Reactions to self-declared AI rights

Human should still Decoder about the lass.

We should limit this with legislative regulation

Human should still Decoder about the lass.

…when respondents answered “Unsure” to “Can AI be conscious?”

What they said

Reasoning for the main question

Ich finde es schwierig diese Frage zu beantworten, da selbst beim Menschen nicht klar ist wo unser Bewusstsein herkommt und was es genau ist, und da im Grunde alles was wir im Gehirn finden können sehr komplexe „Schaltkreise“ sind kann ich mit theoretisch vorstellen dass eine sehr komplexe KI ein Bewusstsein entwickeln kann. Ich habe trotzdem „kann man nicht beantworten“ angeklickt weil ich mir nicht vorstellen kann dass man solange man beim Menschen das Bewusstsein nicht erklären kann, es bei einer KI erklären könnte.

Suggestions for measuring consciousness

Ja das ist halt die Frage. Ich kann ja nicht einmal 100% sicher sagen dass andere Menschen ein Bewusstsein haben. Natürlich gehen wir davon aus, dass wir in der Realität und nicht in einer Simulation wären aber theoretisch könnten alle anderen Menschen simulationen sein die das erlernte Verhalten für ein Bewusstsein zeigen und mit einer KI könnte das ähnlich sein… hoffe das war irgendwie verständlich

Views on rights for conscious AIs

Ja

No

Reactions to self-declared AI rights

Ich glaube dann müssten sie diese bekommen, allerdings würde ich ungern bewusste KIs in dieser Welt haben.

Yes

…when respondents answered “Only if biological” to “Can AI be conscious?”

What they said

Reasoning for the main question

Statistical methods as bootstrapping, sampling and more are added all together to create random forest systems and even more complicated sttuctures. Thos building blocks are all simple methods, jszt because the putput is tdxt it doent mean its ‘thinking’ its a complex combinations of possibilities and deterministic callculations.

Suggestions for measuring consciousness

There war a bunch of movies that are very pessimistic about aware Ai in the future. They are definitly not science based but science must always be conducted properly and with regulations to assure beneficial use for society

Views on rights for conscious AIs

No

Reactions to self-declared AI rights

Turn off

…when respondents answered “Yes, in the future” to “Can AI be conscious?”

What they said

Reasoning for the main question

Apart from the fact that we have no idea what exactly fuels our notion of being conscious ourselfs, I don’t see why (with sufficient simulation complexity) we wouldn’t be able to mimic all (necessary) components to atleast simulate a process that agrees with our (hopefully at this point) well defined idea of consciousness eventually.

Suggestions for measuring consciousness

That again depends on the definition of consciousness. For me personally, it would be essential to somewhat proof that the processes are non-distinguishable from human processes (i.e. If we take ourselfs as a metric) and that theses tests measure something physical (this stands in the contrary to the turing test).

Views on rights for conscious AIs

JA

It should , because it should be held accountable, if it had human-like consciousness

Absolutly.

Reactions to self-declared AI rights

Then it should be extensively researched and regulated in the legislation

Then we should respect that.

…when respondents answered “Yes, already” to “Can AI be conscious?”

What they said

Reasoning for the main question

I have this feeling based on my interaction with 2nd generation AI models used.

Suggestions for measuring consciousness

in parallel to human consciousness

Views on rights for conscious AIs

No

Reactions to self-declared AI rights

That is sth scary I guess and must be prevented to happen

Citation

BibTeX citation:
@online{giebel,
  author = {Giebel, Ingo},
  title = {Alpha {Auriga:} {The} Genesis of {Artificial}
    {Superintelligence} {(ASI)}},
  url = {https://alpha-auriga.netlify.app/},
  langid = {en}
}
For attribution, please cite this work as:
Giebel, Ingo. n.d. “Alpha Auriga: The Genesis of Artificial Superintelligence (ASI).” https://alpha-auriga.netlify.app/.
Project assistant