Topic aside, I just wanted to say I appreciate the discourse here. This level of content and discussion is something I find is sorely missing from the one liners of social media and elsewhere in other topics on the core77 forums as of late.
Reading through the link to the 2010 discussion nxakt linked is amazing. Thanks! The mention of “2017” as a far off (!) future scenario …
I’ll come back for more thoughts on topic but just had to leave that here.
That is where my team in Florida are currently exploring with the technology. Last year we worked off a brief to design a 40’ center console RIB boat for the Florida offshore market. We used traditional design workflows of naval architecture research into digital sketching into 3D CAD modeling and then imaged them in Fusion. From there we utilized AI imaging tech to look for detail directions. There is a lot to consider in a product this complex but we got it down to about 3 weeks per 3D cad concept per designer (including interior designs). I’m currently working with another engineering team to develop further the details of the concepts utilizing current AI imaging tech that has come a long way since last summer. This process step is where the training data is proving to be non-existent we are finding. For obvious reasons, I cannot show anything further due to confidentiality agreements.
The composite image below shows one of the virgin process concepts in 3D CAD (center) and the AI imagery that were created afterwards (periphery) to flesh out possible detail directions which represent the next step in development. Again, this was last summer, so the AI images will look rather primitive.
Hard to separate the style of rendition from the quality of the idea. By that I mean if a mid-level designer had sketched any of these concepts, pen on paper, or Pencil on iPad, they’d probably get minimal looks or desire to continue. This is where the process gets confusing (for me) - unformed and largely irrelevant ideas, rendered magnificently. For some reason I have not found inspiration through this process, granted I’ve spent a total of maybe 30 minutes messing around with it.
Because I have no real opinions of my own, I’ll side with Mr. Cave:
I’m a songwriter who is engaged, at this very moment, in the process of songwriting. It’s a blood and guts business, here at my desk, that requires something of me to initiate the new and fresh idea. It requires my humanness. What that new idea is, I don’t know, but it is out there somewhere, searching for me.
Good interview with Timnit Gebru who used to be co-head of AI ethics at Google.
I feel like the hype of AI is coming from people being impressed with personal first contact with the new tools, chatGPT in particular, and passing that around.
The interview suggests the excitement is just a continuation of previous failed hype cycles such as crypto, nft, metaverse and web3. The hype vector is not genuine excitement, but tech companies pushing AI into legacy media. For AI there wasn’t any evidence given that it would disappoint, just that it is similar to other hype that preceded it.
However, I am just a ‘rando’. 
Well, one of our product lines in oral care accounts for about 15% of the BU’s revenue, so I am interested. What intrigued me about your post was the line “Over 82% of people who brushed with Oral-B SmartSeries experienced a noticeable improvement in their oral health.”
If interested, here is the study - Introducing the Oral-B iO electric toothbrush: next generation oscillating-rotating technology - PMC
I will give it credit that Oral-B did pay for a RCT, not cheap by any means. And in the end, their p-value was < 0.001. Good on them.
But…
Their summary “paper” reads as a sales brochure. When you dive into the individual studies, you see the trials were not on a large scale. Manual brushes showed a 24% improvement, sonic brushes were at 53% improvement over 16.7 bleeding sites. So after use of the products for 8 weeks (kind of short timing and likely has a Hawthorne effect), the iO had 3 bleeding sites, the sonic had 8 bleeding sites and the manual had 13.
And of course the biggest problem is that rotating oscillating brushes have been proven better performers since at least the 90s. How does their AI brush for $500 to compared to a “dumb” RO brush for $50?
/sarcasm tag was missing. Sorry about that.
**Global Economics Analyst **
The Potentially Large Effects of Artificial Intelligence on Economic Growth
Briggs, Kodnani
If generative AI delivers on its promised capabilities, the labor market could face
significant disruption. Using data on occupational tasks in both the US and
Europe, we find that roughly two-thirds of current jobs are exposed to some
degree of AI automation, and that generative AI could substitute up to one-fourth
of current work. Extrapolating our estimates globally suggests that generative AI
could expose the equivalent of 300mn full-time jobs to automation.
Link to the article: Futurism: Goldman Sachs Salivates at AI’s Potential to Mass Fire Workers
Attached PDF of the research.
I did get the sarcasm about AI, no worries. But I do have several patents in the oral care space and hopefully your sarcasm was not to toothbrushes. May not be your thing, and there is nothing wrong with that, but I get geeky about it. 
And the idea of measuring the time on each tooth and where on the tooth is intriguing, but I don’t understand why it takes AI to do that, and Oral-B never explains. But currently I think that is the case for AI, it is AI for AI sake. I’m too dumb to apply it in a meaningful way. I hope someone can give me an idea on how.
I asked chatGPT, “From an industrial designer perspective, discuss how AI could be applied to an electric toothbrush, specifically measuring the time on each tooth and where on the tooth the brushing is taking place”
One possible approach is to integrate sensors in the toothbrush that can detect the movement and pressure of the bristles against the teeth. These sensors could send data to an AI algorithm that analyzes the motion patterns to determine the location and duration of brushing on each tooth.
Another approach could be to use computer vision technology to analyze images of the teeth captured by a camera in the toothbrush. By using machine learning algorithms, the AI system could detect the location and shape of each tooth and track the movement of the brush in real-time. This approach could provide more accurate data on the location of brushing on each tooth.
The data collected by the AI system could be presented to the user in real-time or stored for later analysis. For example, the toothbrush could have a small display that shows a map of the mouth with different colors indicating the areas that have been brushed for a sufficient amount of time. The user could adjust their brushing technique and duration based on the feedback provided by the toothbrush.
Furthermore, the AI system could also provide personalized recommendations based on the user’s brushing habits and dental history. For instance, the AI could detect areas of the mouth that require more attention and suggest specific brushing techniques or additional dental care products.
Me: Make a set of bullet points on how this approach could be tested.
AI could be applied to an electric toothbrush to measure the time and location of each tooth brushing, which could lead to more effective and efficient oral hygiene. Here are some bullet points on how this approach could be tested:
- The first step would be to gather data on typical brushing patterns and behaviors, including the average amount of time spent brushing and the areas of the mouth that tend to be missed.
- Once this data is collected, an AI algorithm could be trained to analyze the motion and pressure patterns of the toothbrush to determine where on the teeth it is being used and for how long.
- The AI could then provide real-time feedback to the user on their brushing habits, alerting them if they are spending too little or too much time on a particular area or if they are applying too much pressure in certain areas.
- The toothbrush could also be equipped with sensors that measure the pressure being applied to the teeth, allowing the AI to further refine its analysis and provide more accurate feedback.
- To test this approach, a group of volunteers could be recruited to use the AI-enabled toothbrush for a set period of time, while another group uses a traditional toothbrush. The two groups could then be compared based on factors such as overall oral hygiene, gum health, and tooth decay rates.
- The AI algorithm could also be tested and refined based on feedback from users, as well as data gathered from a larger sample size over a longer period of time.
- Finally, the toothbrush could be designed with a user-friendly interface that allows users to easily access and understand the feedback provided by the AI. This could include visualizations of brushing patterns and progress over time, as well as suggestions for how to improve brushing habits.
Me: assuming success, present the above in consumer-friendly marketing terms as if in a product brochure
Introducing our latest innovation in oral hygiene: the AI-enabled electric toothbrush. With the power of AI technology, this toothbrush takes brushing to a whole new level, helping you achieve optimal oral health.
Our AI algorithm analyzes the motion and pressure patterns of the toothbrush to provide you with real-time feedback on your brushing habits. The toothbrush measures the time and location of each tooth brushing, alerting you if you are spending too little or too much time on a particular area or if you are applying too much pressure in certain areas.
Imagine having your own personal dental coach in the palm of your hand! The AI-enabled toothbrush is designed to help you improve your brushing habits, leading to better oral hygiene, gum health, and tooth decay rates.
The toothbrush is easy to use and comes equipped with sensors that measure the pressure being applied to the teeth, allowing the AI to provide even more accurate feedback. Our product is user-friendly and provides you with visualizations of brushing patterns and progress over time, as well as suggestions for how to improve brushing habits.
Join the future of oral hygiene with our AI-enabled toothbrush. Experience the power of AI technology for yourself and achieve the best possible oral health!
Confucius say “One man’s sarcasm is another man’s career”. /s (he didn’t really say that)
Consider the Boards a safe space for wisecracks about AI-enabled fitness equipment.
OK. I’m in more in the camp of, that’s just an algorithm, sophisticated, but nonetheless just an algorithm. I suppose we could get into the semantics, but I don’t want to.
I will say that the AI you provided did not take it to the next level of - Does this AI toothbrush actually change behavior? Did Oral-B use AI to create the studies to show outcomes? And since the studies they did were for only 8 weeks, any AI may “think” 8 weeks may be enough time to get an outcome. In my experience you see a Hawthorne effect for at least 12 -16 weeks before you see the inevitable decline in compliance and then you outcomes can go right down the drain.
Also, do any of these AI cite their sources? How do we evaluate their decision-making process? I might be the foremost expert in bullshit, but it is still bullshit.
Geez that’s unreal. Is anyone going to need to write for work anymore? Or just proof and edit, while our brain’s capacities for recognizing original content withers away?
Me: regarding the testing of the toothbrush, would you expect the see the Hawthorne effect in a trial lasting 8 weeks? What would be some psychological methods to reduce the Hawthorne effect?
It is possible to see the Hawthorne effect in a trial lasting 8 weeks, especially if the participants are aware that they are being observed or tested. The Hawthorne effect refers to the tendency for individuals to modify their behavior or performance in response to being observed, which can result in a distortion of the true results of the study.
To reduce the Hawthorne effect, here are some psychological methods that could be used:
- Single-blind or double-blind study design: In a single-blind study, the participants are not aware of which group they are assigned to (e.g. experimental or control group). In a double-blind study, both the participants and the researchers are not aware of which group the participants are assigned to. This can help reduce the Hawthorne effect by minimizing the participants’ awareness of being observed.
- Minimizing the observer effect: The observer effect refers to the tendency for the researcher or observer to unintentionally influence the participants’ behavior or performance. This can be minimized by training the observers on how to maintain a neutral and non-influential stance during the study.
- Minimizing demand characteristics: Demand characteristics refer to the cues or expectations that participants pick up from the study environment or instructions, which can affect their behavior or performance. To minimize demand characteristics, the study instructions and environment should be designed to be as neutral and non-influential as possible.
- Ensuring participant confidentiality: Assuring participants that their data will be kept confidential and that they will not be individually identified can help reduce the Hawthorne effect, as participants may feel more comfortable behaving naturally and honestly.
- Randomizing participants: Randomly assigning participants to different groups or conditions can help reduce the Hawthorne effect by ensuring that any observed changes are due to the experimental manipulation rather than individual differences or expectations.
Me: Can you cite some of the sources that were drawn upon to make the above conclusions?
Sure, here are some sources that were drawn upon to make the above conclusions:
- Stangor, C., & Walinga, J. (2014). Introduction to psychology (1st Canadian ed.). FlatWorld.
- Rosenthal, R. (1966). Experimenter effects in behavioral research. Appleton-Century-Crofts.
- Orne, M. T. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17(11), 776-783.
- Babbie, E. (2013). The practice of social research (13th ed.). Wadsworth.
- Fisher, R. A. (1925). Statistical methods for research workers. Oliver and Boyd.
Naturally, bullshit detection is paramount, But if you spot an error, or want a specific field of research to be referenced, you give the correction or instruction.
Having played with it a little bit, it changes the way I pose questions in all areas. It also changes my search queries, I want more specifics and some inferred meaning instead of thousands of results containing the phrase I searched for.
It contaminates reading discourse because everything becomes suspect. Enough was suspect already.
Thanks again for the input. It seems that AI is driven mostly/entirely by human input at this time. I think it will be much more interesting when it relies on itself.
And for the record, I cite a study from 1925, I lose credibility. Just sayin.
Then you simply set the parameters not to consider input from before 2013.
Me: regarding the testing of the toothbrush, would you expect the see the Hawthorne effect in a trial lasting 8 weeks? What would be some psychological methods to reduce the Hawthorne effect? Please consider only research and papers from the last ten years as resources.
The Hawthorne effect is a phenomenon in which participants modify their behavior in response to the knowledge that they are being observed or studied, leading to changes in the outcome of the study. While it is possible that the Hawthorne effect could be present in a trial lasting 8 weeks, it may be less likely in this context as the participants are likely to be using the toothbrush in their own homes and on their own time.
However, there are several psychological methods that can be used to reduce the potential impact of the Hawthorne effect in such a trial. Here are some examples, based on recent research:
- Use a control group: In order to distinguish between the effects of the toothbrush and the effects of being observed, it may be useful to include a control group of participants who use a regular toothbrush and are not observed or monitored in any way.
- Minimize the impact of the monitoring: Researchers could try to minimize the impact of the monitoring by using unobtrusive methods to collect data, such as sensors embedded in the toothbrush that do not require any active participation from the participant.
- Provide clear instructions: It may be helpful to provide clear instructions to participants on how to use the toothbrush and what is expected of them, in order to reduce any confusion or uncertainty that could lead to changes in behavior.
- Normalize the monitoring: Researchers could attempt to normalize the monitoring by emphasizing that it is a routine part of the study and that all participants will be monitored in the same way.
- Use a placebo group: Researchers could include a placebo group of participants who are given a toothbrush that looks identical to the AI-enabled toothbrush but does not actually collect any data. This could help to control for any placebo effects that may be present.
Overall, while the Hawthorne effect is always a potential concern in any research study, there are several strategies that can be employed to minimize its impact and increase the validity of the study.
Me: cite your sources for the above.
Sure, here are some recent research papers that support the methods for reducing the Hawthorne effect in research studies:
- Harth, N. S., et al. (2018). The Hawthorne effect in direct observation research with physicians and patients. Journal of Evaluation in Clinical Practice, 24(2), 238-242.
- Jenkins, J., & Hodge, C. (2019). Reducing the Hawthorne effect in healthcare improvement projects: A systematic review. BMJ Quality & Safety, 28(1), 54-64.
- O’Brien, K. V., et al. (2020). Understanding and reducing the Hawthorne effect in digital health trials. Journal of the American Medical Informatics Association, 27(7), 1121-1127.
- Peiro, S., et al. (2019). The Hawthorne effect in direct observation research on hand hygiene: A systematic review. Journal of Hospital Infection, 103(3), 245-254.
These papers provide evidence-based recommendations for reducing the potential impact of the Hawthorne effect in research studies, including the use of control groups, unobtrusive data collection methods, clear instructions, normalization of monitoring, and placebo groups.
