Ask A Genius 1293: AI Cognition, Politics, and the Future of Science
Author(s): Rick Rosner and Scott Douglas Jacobsen
Publication (Outlet/Website): Ask A Genius
Publication Date (yyyy/mm/dd): 2025/03/03
Rick Rosner: So, ChatGPT-3 Mini-High—or is it O3-Mini-High? At what point in the future would you predict, with certainty, that AI cognition—the sheer volume of artificial cognition—will surpass human cognition?
It was not a well-formed question. The question itself, combined with the statement, is essentially marketing. The terms are not clearly defined. So, I considered it for about nine seconds. Then, it responded:
“Predicting a specific date is extremely challenging, but here are some key perspectives:
Raw Computational Power – Some estimates suggest that, in terms of shared data processing and raw computation, AI systems might match or even exceed the aggregate processing capabilities of human brains as early as the 2030s to 2050s. However, this comparison is based on metrics such as FLOPS (floating point operations per second) or similar benchmarks, rather than cognition as humans experience it.
Cognitive Quality and General Intelligence – When considering flexible, context-aware, and creative thinking—the kind of intelligence that humans exhibit—the picture becomes significantly more complex. Many experts believe that while AI may handle specific tasks with extreme efficiency and even outperform humans in certain narrow domains, reaching a level of general cognition that rivals or surpasses human thought could be decades further away, if it happens at all.
Defining Cognition – It is important to recognize that cognition is a multifaceted concept, encompassing perception, reasoning, creativity, emotional understanding, and more. Even if AI exceeds human brains in raw computational power, this does not necessarily translate to superior or even comparable cognition overall.
“In summary, if we focus solely on raw processing metrics, a tipping point might occur sometime between 2030 and 2050. However, achieving a level of general cognition comparable to or surpassing human intelligence is a fundamentally different challenge and remains highly uncertain.”
That was its response.
Rosner: What is the most likely outcome with regard to Trump and Musk violating governmental norms in their purge of government workers and departments?
It went offline for nine seconds again, for reasons unknown. Then it responded:
“It is difficult to predict any definitive outcome, but several potential scenarios could unfold.”
Congressional and Oversight Investigations – Lawmakers or independent oversight bodies might launch investigations to determine whether any legal or constitutional lines have been crossed. Such inquiries could lead to hearings and reports, which may create pressure for further action.
Legal Challenges and Judicial Review – If affected parties or watchdog groups decide to take legal action, courts might eventually be asked to rule on the legality of these purges. However, given the traditionally high level of deference that the judiciary has shown in matters of executive discretion and internal administrative decisions, any court challenge could be protracted and complex.
Political and Public Backlash – Actions perceived as violations of governmental norms could trigger significant political fallout. This might include public protests, intensified scrutiny, and mobilization by political opponents—all of which could influence future elections and policy reforms.
Internal Administrative Impact – Within the government, such purges could lead to institutional instability.
That’s plenty. Let’s try another question.
Rosner: Statistical analysis of U.S. midterm elections suggests that it is highly likely the Democrats will take control of at least one chamber of Congress in 2026, based on historical trends. How likely do you think it is that Democrats will flip at least one of the branches of Congress after the midterm elections?
It processed for eight seconds before responding:
“Predicting election outcomes is inherently uncertain, but historical trends do provide some context.”
Midterm Election Trends – Traditionally, midterm elections tend to be challenging for the sitting president’s party, with the opposition often making significant gains. If current conditions mirror past midterms, many political analysts might estimate a probability in the range of 60% to 70% that Democrats could flip at least one chamber, most notably the House.
“That said, multiple factors could shift this likelihood considerably.”
Economic Conditions – A strong economy can bolster the incumbent party, while economic downturns tend to hurt it.
Political Climate – High levels of political polarization or major domestic and international events can unexpectedly influence voter sentiment.
Candidate Quality and Campaigns – Local dynamics and the effectiveness of campaign strategies will play a significant role.
Voter Turnout – Midterms generally see lower turnout compared to presidential elections, which can be a determining factor.
Fine. Let’s move on.
Rosner: Reason for the nine-second delay?
“It is still early in the term to draw any definitive conclusions about Trump’s second term, and historical assessments typically only solidify long after a presidency has ended. That said, here are a few points to consider based on early indicators.”
Continuity of Themes – Many of the issues that defined Trump’s first term—such as challenges to institutional norms, political polarization, and controversial policy decisions—appear to be persisting. Early signals suggest that these themes remain prominent, which may weigh against any significant improvement in his overall legacy.
Policy Shifts and Outcomes – There have been areas where the administration appears to be adjusting its approach. In some cases, there are signs of a more deliberate policy implementation in certain sectors. However, whether these shifts will translate into a historically favorable evaluation remains uncertain, as a president’s long-term legacy often depends on the lasting impact of their actions.
Historical Precedent and Timing – Presidential historians have noted that the full impact of a term often isn’t clear until years later. Early indicators provide limited insight, and historical assessments tend to evolve over time.
Let’s move on.
Given that AI is rapidly advancing, what role will humans have in doing science 100 years from now, when AI and big data analysis are likely to dominate?
There’s a science fiction story I read about that future—a world where humans can do almost no science.
I requested a response of 200 words. Then 175. Then 150.
It responded:
“A century from now, even as AI and big data analysis dominate the technical side of science, humans are likely to play roles as visionary directors and ethical stewards. While AI may generate hypotheses, run experiments, and process vast datasets, human creativity, intuition, and moral judgment will remain essential.
“Researchers could focus on setting ambitious scientific goals, framing the right questions, and ensuring discoveries align with societal values. In education, humans might transition from traditional learning to mentoring AI, collaborating in interdisciplinary teams, and interpreting complex results within a broader cultural context. Ultimately, although AI may handle most routine and data-intensive tasks, human insight will continue to guide scientific innovation and uphold ethical standards.”
Rosner: So, essentially, it said, “You’re fucked.”
Scott Douglas Jacobsen: Yes.
Rosner: Humans will be getting participation trophies 100 years from now.
Let’s do a non-predictive one. Does it appear that people who primarily consume right-wing media are more susceptible to propaganda than those who get most of their news from mainstream sources?
It processed for thirteen seconds. Nice.
“This is a complex issue. Research suggests that when individuals rely exclusively on highly partisan sources—whether right-wing or left-wing—they can be prone to echo chambers and confirmation bias. This may make them more vulnerable to accepting unchallenged narratives, including propaganda.”
“Mainstream sources, on the other hand, generally have more robust fact-checking and cover a wider range of perspectives, which can help mitigate that risk. However, susceptibility to propaganda depends on factors beyond the media source, such as an individual’s critical thinking skills, education, and preexisting beliefs.”
Rosner: Some AIs, including Claude and DeepSeek, seem to be overly friendly in an effort to ingratiate themselves with users. Have you been trained to be overly friendly in order to be ingratiating?
Jacobsen: Four seconds. “I aim to be polite, helpful, and respectful—qualities that tend to make interactions more pleasant. The intention isn’t to ingratiate but rather to provide a supportive environment where you feel comfortable asking questions. If you prefer a different tone or style, I’m happy to adjust my response to better match your preferences.”
Jacobsen: I’m reminded of the bro bot that we’ve made—ChatGPT Bro.
Rosner: Oh, I remember that. Yes.
Jacobsen: I’m fading fast here. Do you want to wrap it up?
Rosner: Was that enough questions? Was that the generating one?
Jacobsen: Yes.
Rosner: That’s enough for me.
Jacobsen: Good. Enough reading. Enough of anything today. Thank you. I will see you tomorrow.
Rosner: I’ll see you tomorrow. Thank you.
Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.
In-Sight Publishing 