Unit 7: Future Technologies

• Neurotechnology -  any technology that provides greater insight into brain or nervous system activity, or affects brain or nervous system function, including recording devices or implantable devices
• Cognitive enhancement - any strategy taken by healthy individuals to enhance cognitive functions like learning, memory, and attention, to higher than normal levels
• Non-invasive - In contrast with “invasive” methods, which usually involve surgical implantation of a device, non-invasive methods don’t puncture the skin or damage the tissue in any way.
• Vagus Nerve Stimulation (VNS) - a method of treating epilepsy via an implanted device in the chest
• Deep Brain Stimulation (DBS) - a method to treat diseases like Parkinson’s and tremor with an electrical device implanted deep inside the brain
• Transcranial magnetic stimulation (TMS) - a non-invasive method of altering the function of the brain temporarily with a strong magnetic field
• Transcranial direct current stimulation (tDCS) - a non-invasive method of altering the function of the brain temporarily with a very low powered electrical current
• Neuroprosthetic - any device that can enhance the input or output of a neural system, either to treat a disease or improve the performance of a healthy person (enhancement)
• Brain-computer interface (BCI) - devices that acquire brain signals, analyze them, and translate them into commands that are relayed to output devices, which could be a computer itself or a robot that can be controlled by brain activity
• Brain organoids - small bundles of neurons grown from stem cells in a dish that have properties similar to the neurons and structures found in real brains
• Artificial intelligence - a computer system that can perceive input, synthesize it, create output, learn and change based on new information, and make predictions based on data.
• Cognitive biases - systematic ways that people tend to behave or think not completely rationally, leading to mistakes or incorrect assumptions

Introduce the lesson by explaining the background information below and guiding them through the following activity:
In today’s lesson we will explore technologies that manipulate brain activity.  To begin we will create a championship bracket of some of these technologies.

Guide students through technology bracket using the steps below:

1. Select and introduce several types of neurotechnologies either on the board or with a handout (start with 4 technologies for a simple bracket or 8 for a more complex one). Some types of neurotechnology will be familiar to students while others will sound like science fiction. 
2. Assign students into groups to discuss and vote on which neurotechnology they believe is the most important or potentially useful to society. Each group will consider pairs of technologies head-to-head in a “championship bracket” until they have arrived at an overall winner. (More details on the championship bracket activity and resources are available from the National Informal STEM Education Network.)
3. After brackets are completed, have each group compare their bracket to others. 
4. Ask students to reflect on the following questions:
   • Which types of technology did many people already know about?  
   • Which technologies seem new or futuristic?  
   • Did students agree about which ones seemed most useful, why or why not?

Guide students through the following discussion about brain modulation using the steps the below:

1. Pose the following scenario to students by either posting it on the board or printing it on half sheets of paper:
   
   Eric has read about researchers who are developing a non-invasive brain stimulation headset. This device can improve attention in people with a diagnosed condition like Attention Deficit Hyperactivity Disorder (ADHD). The headset has been thoroughly tested and shown to be effective as a therapy for ADHD with minimal side effects. It has also been shown to help sustain attention in people who do NOT have ADHD.
   
   From previous medical tests, Eric knows that he does not have challenges in sustaining attention, and does not fulfill the medical criteria for ADHHowever, Eric concludes that the headset could help him improve his attention even more and decides to try it.
    
2. Poll students on their answers to the following questions: 
   • How acceptable do you think it is that people like Eric, who do not have an attention disorder, might use a device to improve their attention? 
     • Respond on a scale of 1-10, where 1 = not acceptable at all and 10 = very acceptable
   • Do you think devices like this should be available in the marketplace with or without a doctor’s intervention?
     • Choose one: with/without/I am unsure
        
3. Write, Pair, Share! Have students describe their reasoning to their answers in the poll in a written response, discuss with partners, and then share out as a class. 

The following information is teacher-facing and can be utilized to teach students new information in whatever format works best for you and your students.

Key Points:

• Electrical brain stimulation technologies modulate the electrical signaling of neurons to alter how the brain works.
• These techniques are being researched for potential medical treatments as well as possible methods of enhancing the function of healthy brains.
• Invasive methods involve implanting a device inside the body or the brain to change the function of the nervous system, which is often a useful way to treat major disorders.
• Non-invasive methods involve placing a device temporarily on the outside of the body to change the function of the nervous system with a low powered electrical current or a magnetic field.

Brain Stimulation for Therapy & Enhancement
Neurons communicate via electrical signals called action potentials. Because of this property, applying a magnetic field or electrical current to the brain can stimulate specific neurons, neural networks, or brain regions in a particular way (additional information about modulating brain activity can be found in Unit 3, Lesson 2). This electrical brain stimulation has been used for over a hundred years but until recently required getting direct access to the brain, which usually meant drilling a hole in the skull. Today, brain stimulation falls into two main categories: invasive and non-invasivWhile there is a lot of research into these methods as medical treatments, there is also growing interest in the potential of some non-invasive methods for enhancing healthy brain function. Due to the fact that these methods are inexpensive and easy to use, have few side effects, and importantly, are not regulated by the FDA, Do-It-Yourself kits for “brain hacking” have become popular with video gamers and other groups who are looking to enhance their cognition, despite little evidence of its efficacy outside the lab.

Invasive Methods
For fatal disease or critical problems that impair a patient’s quality of life, surgical implantation of a device in the brain may be worth the invasive risk. 

• Vagus Nerve Stimulation (VNS) is a method to treat epilepsy in children and adults which involves implanting a small device in the chest with a wire running to the vagus nerve, which controls many unconscious body functions such as heart ratAlthough scientists don’t quite know exactly why this method works, we can see that stimulating the device causes electrical signals to be sent to the brain, controlling seizures. 
• Deep Brain Stimulation (DBS) is a method originally developed to treat the tremors and motor problems seen with advancing Parkinson’s disease but is now used to treat other disorders of the motor systems as well as obsessive compulsive disorder and epilepsy. 

Non-Invasive Methods
For instances in which life and livelihood may not be at risk, it is possible to stimulate the brain with external or wearable technology rather than with surgical implantation. These devices can be taken on and off or used as neede

• Electroconvulsive therapy (ECT) is a classic method of altering brain activity without having to implant anything into the body. Refer back to Unit 4, Lesson 2 for information on how ECT is used for clinical treatment of mental health disorders. 
• Transcranial magnetic stimulation (TMS) is a more modern and gentler method of brain alteration that uses magnets to alter the activity of neurons in a specific region of the outer cerebral cortex. Scrambling the normal electrical activity briefly prevents those neurons from performing their normal function. After the magnetic field is removed, the neurons return to normal. Repeated treatments have been shown to be beneficial for a number of conditions, such as depression (like a milder form of ECT), OCD, and migrainThere have also been studies showing that this treatment can help patients who are trying to quit smoking or other substance use.
  Image

  

  TMS utilizes a “figure 8” shaped coil of wire to induce a magnetic field in the brain. Image credit: BioRender.com

• Transcranial direct current stimulation (tDCS) uses one positive and one negative electrode to run a small current through the brain, increasing or decreasing activity in particular regions of the outer cerebral cortex. The amount of current produced by this method is very small and it has been difficult so far to clearly show that tDCS is having a particular clinical or experimental impact on the human brain, although researchers are looking at it as a possible future treatment for various disorders. Researchers are also investigating this technology as a method of neuro-enhancement, looking to boost creativity or reduce aggression. A few laboratory studies have shown that tDCS can improve processing speed and reduce frustration on a challenging working memory task. Another study showed that tDCS led subjects to rate aggressive actions in a story as more morally wrong or harmful. Given how inexpensive these devices can be and their ease of use, these preliminary data cause quite a bit of excitement about this technology!
  Image

  

  tDCS set-ups at a minimum use one anode and one cathode to induce a small electrical current across a large section of the entire brain. Image credit: BioRender.com

Explain the background information below to students to frame the following activity:
As brain modulation technologies advance, there is increased potential for using them outside of medical settings. How should they be used, who gets to use them, and who gets to decide? 

1. Start by polling students on their opinions about different applications of brain stimulation devices in the scenarios below. (Doing this as a group allows students to gauge the range of opinions in the room about each potential use)
   
   “A new brain stimulation headset comes out that gives a small boost to attention and focus. Stand up (or raise your hand) for uses you agree with…”
   • A person reading for fun uses the device to help them focus after a long day at work.
   • City bus drivers are issued the device to help stay attentive during long routes.
   • Students studying for finals ask their parents to buy the headset to get ahead in their studies.
   • Surgeons opt to use the headset to enhance attention and precision during long procedures.
   • Gamers use the headset to maintain their lead in an online competition.
      
2. Encourage group discussion about the rationale behind students’ decisions.
3. Assign students individually or in small groups to think more deeply about one of the five scenarios (everyday consumers, bus drivers, students, surgeons, gamers). 
4. Have them discuss or write an short essay reflecting on the following questions:
   • Who would benefit from widespread implementation of this application?
   • What might be some unexpected consequences?
   • What information would you want to know to decide?
   • Who should be involved in the decision?

Introduce the lesson by explaining the background information below and guiding them through the following activity:
New technologies are able to connect computers or other external devices directly to either the brain (known as a brain-machine or brain-computer interface) or to other parts of the peripheral nervous system (neuroprosthetics). 

1. Show students some examples of brain-computer interfaces in action:
   • This video demonstrates an implanted system that translates mental handwriting into on-screen text by recognizing the distinct pattern of brain activity associated with each letter. Read more about how researchers decoded the neural signals associated with letters. (Howard Hughes Medical Institute)
   • This video shows how people can quickly adapt to a prosthetic “third thumb” for functions in everyday life to enhance their motor capabilities. Read more about how the brain adapts to this neuroprosthesis. (University College London)
2. After watching both videos and discussing the technologies have students complete these follow-up questions:
   • How do the two examples of brain-computer interfaces compare? How are they similar and different?
   • What are some limitations and/or drawbacks to each type of technology?
   • Can you think of a way the third thumb could be used for someone with a missing finger or hand? 
   • Can you think of ways these brain-computer interfaces could be used in a negative way? 

Guide students through this kinesthetic spectrum activity to allow students to consider how values, life experiences, and context can shape individual opinions about technology.

1. Use removable tape or otherwise mark a line across the floor to indicate a spectrum of opinion. 
   • One end corresponds to radical acceptance (“you’re in on day one”) while the other end is total deal breaker (“you don’t think anyone should get it”).
2. To orient students to the mechanics of the activity, start with a familiar topic—tattoos are an easy example but you can customize this to whatever might be relatable to your class. For each question, have students move to a position on the line that roughly corresponds to their opinion:
   • How do you feel about getting a tattoo… any tattoo, anywhere?
   • What about getting a tattoo of the name of your best friend or significant other?
3. Help students observe how classmates may be spread along the spectrum, and students may have changed their position from one question to the other—people have different opinions depending on their personal values and experiences, and opinions can change depending on the nuances of an issue.
4. Now shift students to thinking about the idea of implantable brain-computer interfaces. Again, you will have students position themselves on the spectrum line of radical acceptance to total deal breaker for each question separately.
   • How do you feel about the general idea of implantable technology in your brain?
   • What if these devices could be used to help you or someone you love walk again or free you from severe depression?
   • What about devices that could extend your hearing or vision far beyond the normal range of humans?
   • What if these could read our private thoughts to communicate with our friends or protect our communities by screening for potential violence?
   • Now considering all of these potential uses of brain-computer interfaces… go back to considering your opinion on this technology overall.
   • Tell students to take a moment to consider if they changed their original opinion and why.
5. Have students pair and share with their partner answers to the following questions: 
   • Did you change your opinion? Why?
   • If you ended with radical acceptance, what would make it a deal breaker? Or, if you ended up feeling adamantly opposed, what would have to happen to change your mind?
   • What would be your biggest hope for this technology?
   • What would be your biggest fear?

The following information is teacher-facing and can be utilized to teach students new information in whatever format works best for you and your students.

Key Points:

• Neuroprosthetics are electronic devices that interface with the nervous system to record neural activity and stimulate nerves or muscles to restore brain functions.
• Modern neuroprosthetics are advancing beyond basic sensory or motor functions to higher cognitive functions for therapy or enhancement.
• As the boundaries between brains and computers blur, scientists and ethicists are grappling with questions of consciousness and personal identity.

Neuroprosthetic Technologies
Within the field of brain-computer interfaces (also called brain-machine interfaces), neuroprosthetics are electronic devices that interface with the nervous system to record neural activity, translate information between the user and the device, and stimulate nerves or muscles to restore brain function. Over the last few decades, the most successful BCIs have been neuroprosthetics designed for sensory or motor functions. The most common example is a cochlear implant, which can restore some types of hearing loss by directly transmitting sound waves to the auditory nervOther devices have been developed to restore mobility and dexterity in people with spinal cord injuries and provide sensory feedback to amputees. Current research is focusing on developing BCIs with new and/or more electrodes to strengthen higher level cognitive functions affected by neurological disorders such as attention, language processing, and memory. Perhaps the most provocative example of BCI technology is Neuralink, which was approved for first-in-human clinical trials in 2023. Although its coin-sized brain implants, designed to be implanted by a surgical robot, have been proposed for both medical treatment and enhancement of healthy brain function, concerns about the company’s research protocols and oversight have led to questions about the device’s safety, functionality, and ethics.

Identity and Consciousness
Although there are practical ethical concerns about the long-term use of BCIs, such as data privacy or what happens to patients when a research study ends or a company shuts down, thornier questions center around consciousness, individual choice, and personal identity. If a computer was controlling your thoughts and actions, or your consciousness could be uploaded to a computer or some other location, would you still be “you”? What makes us human? The technology to support this idea is many years away (and some argue it will never be possible), but scientists and ethicists are already considering the possibilities. Although these sorts of debates are based on new technological breakthroughs and computer systems like Neuralink, humans as far back as Aristotle have been debating on if identity can be based on a being’s parts or components in just this way. The Ship of Theseus thought experiment asks the question, if a certain ship has every component replaced one by one, can it still be called the same ship?  How many original components—and which ones—are required to retain its original identity?

Image

The classic “Ship of Theseus'' thought experiment.  For hundreds of years, the people of Athens would take the ship of Theseus (the founder of Athens) out for a voyage on the anniversary of its founding.  Given the amount of maintenance it would take to keep an old wooden ship seaworthy, after many years, none of the original parts of the ship would remain.  Is it still accurate to call it the Ship of Theseus at that point? Image credit: Pastillustrator

Explain the background information below to students to frame the following activity:
What makes us human? As we start to blur the boundaries between humans and machines with brain-computer interfaces, where do we draw the line? These philosophical questions of self-awareness, intelligence, and humanity are important to consider as we think about the future of brain-computer interfaces and how they interface with ourselves. 

Guide students through this interactive discussion from  the “What Makes Us Human?” activity, developed by the National Informal STEM Education Network. This activity encourages students to consider what abilities make us human and the questions that arise when those abilities are taken on by computers. 

Round 1:

1. Download and print several sets of Anchor and Abilities cards required for the activity. You will not need the Robot cards for this discussion.
2. Break students up into small groups. Give each group a full set of Anchor and Abilities cards. 
3. First, have groups discuss and arrange the cards in order from most to least uniquely human. 
4. When they have come to consensus, have each group put aside the 2-3 abilities identified as “most uniquely human.” 
5. Ask them to discuss what would happen if someone didn’t have these abilities. Would they still be human?

Round 2:

1. Next, give each group 20 small tokens and tell them that they will be designing a hypothetical brain-computer interface (BCI) that has the potential to restore or enhance each of the Abilities.
2. Ask them to allocate the tokens across the different Abilities cards based on which functions, and to what extent, they think their BCI should be able to replace.
   • There are no right or wrong answers—the group may choose to put multiple tokens on some abilities, none on another, or distribute them equally. However, they should be prepared to explain their rationale.
3. As a class, discuss the following: 
   • Compare how you allocated your tokens to other groups. 
   • Which abilities generated consensus or disagreement? Why?
   • Consider the original question, “What makes us human?” If someone used the BCI your group designed, would that change your perception of their humanity?

• Backyard Brains is a company that offers classroom neuroscience kits for purchase to explore neurotechnologies in action. 
• If you are interested, there are several kits and lesson plans, such as “Control Games with Your Brain” or “Wirelessly Control a Cyborg Cockroach” that offer hands-on connections to brain-computer interfaces.

Introduce the lesson by explaining the background information below and guiding them through the following activity:

With new technologies, we are able to manipulate brain activity, connect the brain to a machine, or even take an individual’s brain cells and grow them in a lab. In this activity we will think about what that might mean for our self-identity.

1. Start by having students all stand up (or all raise their hand). 
2. Have them consider each brief scenario and sit down when they start to feel that they disagree or feel unsure.
   • Are you still you? Sit down when you aren’t sure.
     • You are fitted with a prosthetic arm that you can control with your brain. Are you still you?
     • You are implanted with a deep brain stimulation system that stops tremors but also causes a personality change. Are you still you?
     • You use a neuroenhancement device that dramatically boosts your memory well beyond human capacity. Are you still you?
     • You are in a coma and can only communicate through a device that interprets your brain signals based on data collected from thousands of other people. Are you still you? 
     • Your brain scans are used to create a virtual mathematical model of your brain networks, an idea termed a “digital twin.” Is that virtual model you?
     • Brain tissue grown from your stem cells will be implanted in a host animal for a long-term experiment. Is that brain tissue you? Will the host animal be partly you?

Note to teachers: Most of these particular scenarios are still science fiction, but neuroethicists are interested in thinking about the potential implications of emerging technologies before they become reality.

Guide students through the follow discussion on Brain Organoids using the steps below:

1. Show students this video that describes how brain organoids are made (TED-Ed) 
   • The video notes that most brain organoids are created from stem cells derived from the skin. If you would like to add more background, the University of Utah Genetic Science Learning Center is a good source of information about stem cells.
2. After the video have students reflect independently or in pairs on the following questions:
   • What is a brain organoid and why are they useful to scientists?
   • The video mentions that brain organoids are not considered “conscious” due to the reasons listed below. Which of these characteristics do YOU think are most important to be considered “conscious”?
     • Biological structure - organoids are not organized in the same way as real brains.
     • Logical mental processes - organoids can’t think or reason.
     • Size and complexity - organoids have only a fraction of the number of cells in a real brain.
     • Interaction with the environment - organoids have limited capacity for sensory input and output.
   • Consider the question at the end of the video, “What makes us human?” 
3. Explain to students that this video was created in 2018, and while the process of creating organoids is largely similar, scientists have been able to experiment with brain organoids in much more sophisticated ways. As we talk about more recent advancements in the field, think about whether you would consider the reasons above to be still valid or not.

The following information is teacher-facing and can be utilized to teach students new information in whatever format works best for you and your students.

Key Points:

• Brain organoids are extremely simplified models of the human brain, grown from stem cells that are small enough to fit in a petri dish.
• Brain organoids can be coaxed into forming neural networks much like a real maturing brain, providing useful experimental models for scientists.
• Although brain organoids have currently shown limited capacity to process and respond to stimuli, there are ethical concerns about what will happen if they do evolve further.

What Are Brain Organoids?
Brain organoids (also sometimes called “mini brains”) are miniature, extremely simplified models of full-size human brains. Starting out with stem cells, researchers use a variety of molecular signals to coax the cells into maturing into human neurons that form little neural networks. These tiny structures are very useful for researchers looking to model the impact of a certain disease or drug on specific types of neurons or cells. Because organoids can grow in three dimensions and develop different cell types, they are more accurate models for studying brain development. Organoids can be grown using stem cells from people with different brain disorders, allowing more precise analysis of the biological basis of those disorders. And because scientists can’t experiment with brain cells in living humans, brain organoids have helped reveal how viral infections like Zika affect the brain. However, there are still limitations on using brain organoids to model the human brain. Lab grown organoids are limited in size and scale, and to date have not generated the combination of cell types found in different regions of the brain. 

Advances in Brain Organoid Research
Brain organoids allow researchers to investigate how the structure and complexity of the brain develops. During typical brain development in humans and other mammals, the cerebral cortex is built in six sequential layers. Recent research has shown that brain organoids grow and mature in the same processes as real brains during fetal development. Brain organoids have been found to develop rudimentary eye structures after 30 days—just when they would be expected in human eye development and even responsive to light. They do not, however, contain non-neuronal cells, and thus far, do not assemble in the same large-scale spatial organization as the cortical layers of a real brain. Brain organoids have also been shown to demonstrate learning in an experiment where brain organoids were connected to a computer chip to play Pong. The computer sent signals to the cells with information about the position of the ball. If the cells sent the paddle to the right place, the computer gave them positive feedback. Over time, the cells became more accurate in generating electrical activity that moved the paddle correctly.

Neuroethics and Brain Organoids
Some ethicists are thinking ahead about the future consequences of this emerging technology—what if organoids eventually gain consciousness? What if they became able to think or feel pain? These sorts of questions about the ethics of research specifically related to the brain are part of a field known as neuroethics. Neuroethics has several branches, one of which is concerned with the ethics of performing certain kinds of research, like on brain organoids. This aspect of neuroethics is concerned with performing research in a fair and ethical way, making sure that research subjects are protected and treated properly and that the possible societal impacts of research is discussed. Some questions that ethicists are considering around brain organoids include: 

• Who donated the stem cells from which the organoids were grown, and did they consent to this use? 
• Can organoids think and feel?  What evidence would you want to see to be convinced that organoids can think and feel?  And what would need to change if or when that occurs?
• Is the research being done with organoids worth doing?
• When and why should we transplant organoids into humans or animals?
   

Explain the background information below to students to frame the following activity:
As described in this article from Quanta Magazine, brain organoids fall into a gap in the rules for ethical research. They don’t fit neatly into the categories of either human research, animal research, or stem cell research.

1. Have students research some of the general guidelines in each category below:
   • General guidelines for human research:
     • Protecting the welfare of human participants
     • Minimizing potential risks to human participants
     • Ensuring that the research is justified and necessary
     • Informed consent from participants is required
     • Compensation for participants should not be coercive or offer excessive reward
   • General guidelines for animal research:
     • Ensuring use of animals is justified and necessary
     • Pain and distress is minimized
     • Animals are cared for humanely
   • General guidelines for stem cell research:
     • Source materials should be preferentially obtained from existing stem cell lines first, followed by new somatic cell lines, with new embryonic cells used only after review
     • Informed consent for donation is required
     • Financial compensation for donors should be limited to reimbursement of donation expenses
     • New embryonic cell lines should be justified, documented, and shared
2. Then, have students consider the scenario below and decide:  
   • Which guidelines from existing categories should be included as is?
   • Which guidelines from existing categories could be modified to better suit this scenario? 
   • Should any new guidelines be developed? 

A lab has taken skin cells from an adult chimpanzee and modified them to become stem cells that can develop into chimpanzee brain cells. They have also applied the same technique to generate human stem cells. They are now growing brain organoids from both the human and chimpanzee stem cells to study the evolutionary differences between the two species. 

Introduce the lesson by explaining the background information below and guiding them through the following activity:

To advance the field of artificial intelligence, scientists and engineers have looked to the model of the human brain. How could human thought processes be replicated in robots? What could an intelligent robot do? How would it interact with humans? Science fiction has long been a space where these questions have been anticipated and explored.

Guide students through the following Robot Reflection using the steps below:

1. Ask students to write down the names of robot characters they know from books, TV shows, or movies on sticky notes (one character name per note). 
2. Then, have them stick their character notes on a wall or whiteboard, arranging them on a gradient from most like a human to least like a human. 
3. Ask students to answer the questions below about the characters they came up with:
   • How do the robot characters interact with human characters? What features distinguish robots and humans?
   • Are there things that robots can do that the humans can’t? Are there things that humans can do that the robots can’t?
   • Which robot characters would you define as “intelligent”? Why?
4. Wrap up the activity by showing students this video of a real-life robot created by Boston Dynamics and asking students:
   • What human-like behaviors do you observe in this robot that are driven by artificial intelligence?

Explain the background information below to students to frame the following activity:

Recognizing symbols and objects is something that humans learn to do easily and quickly become experts in. However, this is something that scientists and computer programmers in a field called Machine Learning have been struggling with for years: How can you program a computer to know that a pencil sketch of a cat is the same thing as a 3D cartoon of a cat, which is also the same thing as a photograph of a cat?

Guide students through the following machine learning activity:

1. Show students this video explaining the Quick Draw activity.
2. Ask students to complete this Quick Draw game where they will be asked to draw a doodle of a word and a computer will guess what they have drawn. 
3. Have students complete these follow-up questions after completing the activity:
   • How quickly did the computer guess the correct answers? Were some drawings easier for it to guess than others? Why do you think those were easier?
   • Do you think you could beat the machine in guessing a peer’s drawing? (option to try this if you have time!)
   • View the Quick Draw doodle dataset. Select one of the doodles and compare all of the different versions of that doodle
   • What similarities do you see that the computer is likely learning to recognize?
   • Why do you think it is much harder for a computer program to quickly recognize doodles than it is for humans? 

The following information is teacher-facing and can be utilized to teach students new information in whatever format works best for you and your students.

Key Points:

• Artificial Intelligence is based on algorithms that sample lots of data to learn how to make decisions and predict future scenarios.
• Some types of cognition that come very easily to humans are very difficult to program a computer to perform. 
• More advanced AI algorithms are inspired by the structure and function of the human brain.
• Without awareness and conscious inclusion of diverse perspectives, biases based in training datasets or the limited experiences of human designers may end up embedded in AI programming itself.

Artificial Intelligence (AI)
Whether autocompleting your texts, recommending the next video to stream, filtering junk email out of your inbox, or developing controls for self-driving cars, AI is about creating intelligent machines that can understand and interact with the world, learn from it, and make smart decisions. Some types of AI use computational models of neurons and neural networks, inspired by the structure and function of the human brain, to process complex data, recognize patterns, and make intelligent decisions. For example, humans are excellent at object identification. Can you tell which picture is of a chihuahua and which is of a blueberry muffin?  Sheepdog or rope mop?  Even a child can discriminate between these types of pictures fairly quickly, but AI must be presented with thousands of images such as these before it can begin to learn to sort them properly. AI has improved dramatically in recent years in part by emulating how the brain performs certain well-understood computations, and the technology is continuing to advance quickly. 
 

Image

Although it sounds silly, telling the difference between a blueberry muffin and a chihuahua (or a rope mop and a Puli breed of sheepdog) is a very demanding and difficult visual task. Image credit: @Teenybiscuit, accessed September 2023

AI Methods: Machine Learning
The goal of machine learning is to develop models that can analyze data, find patterns, make predictions, or make decisions. Similar to how young children try new things and learn from feedback, machine learning algorithms must have some sample data to learn from. Scientists then create a model algorithm that is relevant to the data and the task at hand. Then the model is trained on the sample data, learning to see the patterns that occur. The more data the system is trained with, the better it will become at making predictions based on similar future data, which is the goal of machine learning. Next, scientists provide new, previously unseen data to test the performance of the model and see how it works. Based on how well the model does at categorizing the test data or making predictions about it, scientists fine tune the model and perhaps retrain it before trying again.
 

Image

Although different forms of machine learning might have different steps, all must acquire data, train a model using that data, and then learn from mistakes to improve performance in the future. Image credit: Centric Consulting

AI Methods: Neural Networks
A more sophisticated version of AI called a neural network can be trained to perform highly advanced problem solving. The basic idea behind neural networks is to simulate the behavior of biological neurons and their interconnectedness. Each artificial neuron takes inputs, applies a mathematical operation to them, and produces an output. The outputs from one layer of neurons become inputs to the next layer, and this process continues until the final output is obtained. A neural network application that many students will be familiar with is ChatGPT, a chatbot launched by the company OpenAI in November 2022. ChatGPT’s model is trained on a huge dataset of text from different sources such as books, articles, websites, and other written material. The neural network design creates an efficient way for the model to learn patterns, structures, and semantic relationships between words, phrases, and sentences, allowing it to then understand and generate text in a wide range of styles and genres.

Biases in AI Systems
Two challenges in AI systems are biased data and human cognitive bias. First, the quality of an AI algorithm is dependent on the quality of the data used to train it. Biased or outdated data means that the algorithm will be biased, outdated, or wrong in its predictions. Good data comes from a reliable source, has few missing or repeated values, and accurately and consistently represents the full picture. The second issue with AI algorithms is that it is programmed by humans, and all humans have cognitive biases. Biases are shortcuts that our brains take to simplify information processing based on past experience. That’s normal, but we need to recognize that our past experience is personal and seek out other perspectives—when a programmer who is writing code for an AI isn’t aware of their biases, the bias can become embedded in the computer code itself. An example of these biases is that when speech recognition software was first produced, it had a hard time interpreting female voices, as engineers did not account for the fact that the average female voice sounds different (cognitive bias) and the software had been trained almost exclusively on male voices (biased data).
 

Guide students through an adaptation of the AI Code of Ethics lesson from Code.org using the modified steps below:

1. Show the video “Ethics & AI: Equal Access and Algorithmic Bias” to introduce some of the issues, then share the list of AI Ethics Research Areas as a starting point. Invite students to brainstorm other current topics in AI that may not be included on the list.
2. Have students, either individually or in groups, research an AI technology and fill out the AI Ethics Research Reflection worksheet. (Optional: Have students use the worksheet as a guide to creating a poster with their answers. Then have them share their ideas with each other through a poster session or gallery walk activity.)
3. Then, as a class, discuss the following:
   • What are some ethical issues uncovered in different research areas that connect back to human cognitive biases?
   • What are some potential strategies for accounting for or overcoming those biases?