Read "The Moral Imperative to Teach AI" to see how educators will determine whether GenAI becomes pernicious or a panacea:

Each year without proper AI instruction is another generation paying the price through diminished cognitive development. But Educators face a challenge: how do you teach something you haven't been trained in yourself?.

It will take decades to acquire data on the impact of GenAI on long-term cognition and dementia. But, let’s be honest, we know some of the consequences:. Reduced cognitive utilization → long-term cognitive decline → increased risk of dementia.

All of this is compounded by the fact that AI-generated explanations are compelling by design. So compelling that people trust AI reasoning over their own—even when the AI is wrong and they are correct.

For professionals, using AI means shifting from being the creator to the reviewer. The problem is you still think you retain the ability to create. For students, it’s similar but worse. They miss out on the opportunity to effectively judge the boundaries of their knowledge.

The evidence is already clear: AI affects learning, reasoning, critical thinking, creativity, and metacognition—our ability to regulate our own thought processes. How we use this tool defines how it changes our cognition.

When you use GPS constantly, your spatial cognition weakens. When you rely on Google, you remember how to find info rather than the info itself. AI takes this adaptation to a new level because it changes how we think.

What if we replaced "How do we keep AI out of the classroom?" with "How do we teach AI as a tool rather than a crutch?".

Check out the full post exploring the future of biological research, automation, and the hidden forces shaping our experiments here:

Automating cell culture isn't about making science easier - it's about making it better. By acknowledging rather than denying natural selection in our labs, we're opening new frontiers in biological research.

As labs modernize and bring in automation, there's an opportunity to do better. We need to rethink cell culture from the ground up, just like the alarm clock reinvented wake-up calls. Enter: self-driving cell culture systems.

This problem has been known for over 50 years, yet most labs still avoid regular cell verification. Some argue time and cost, but fundamentally it's a workflow problem.

In part because of the mechanisms of selection at play, 5-46% of cell lines in research labs are misidentified. Even samples from originating labs are wrong 18% of the time. We're often not working with what we think we are.

Biologists face a paradox: they want to minimize variability to more easily analyze their results, BUT they are studying systems that evolved to have high variability as an evolutionary survival strategy.

So point 1 - natural selection and survival of the fittest takes place EVERYWHERE, including experimental labs.

Take HeLa cells as an example. Did you know these immortal cells are so good at surviving that they literally invade other cell lines?. One study found that 67% of misidentified cells had been overtaken by HeLa cells.

High school bio told you about the peppered moth and it's changes due to pollution during the industrial revolution. Lab specimens are doing the same thing under our noses, but the consequences for research are far more complicated.

Natural selection doesn't just happen in nature - it's actively shaping laboratory experiments (and sometimes undermines them).

Prediction 3: New statistical approaches will .1) attempt to quantify and measure the probability of validity .2) assess the resiliency of the proposed model .3) assess the resiliency of the biological system .4) result in numerical-based assessments of modulatory impact.

Prediction 2: The structure of hypotheses will change, rejecting the model of a single null and alternate hypotheses pair searching for causality in favor of .1) exploring the number of alternate hypotheses explored .2) increased focus and sensitivity to modulatory effects.