Intent helps people eat better and simplify their lives. Figuring out what to eat is a common struggle of modern life - we solve this. Our app makes personalized meal plans tailored to your goals, preferences, constraints, and routine. We're rapidly growing, people love our app, and we're funded by top investors.
We're looking to hire software engineers to help us craft intuitive user experiences for our native iOS app, and to design and scale our backend services. Creating personalized meal plans is a non-convex optimization problem, so the architecture and technical constraints we use are especially important. Our stack includes Swift, Node, Python, Postgres, and Redis. You'll play an integral role in building and scaling the product, and your compensation will reflect this.
Feel free to reach out directly, will [at] intentapp.com
This is an insightful point, and has been demonstrated in several studies [1] - limiting variety simplifies and bounds the food decisions you have to regularly make.
Certainly abuse and addiction management has improved since the 1960s - but companies have struggled to get more stimulants to market because of other side effects as well (e.g. fenfluramine was quite effective, but withdrawn in the 1990s since it was found to cause heart valve disease / pulmonary hypertension). At this point, broader adoption of medications seems to be blocked by a combination of physicians choosing not to prescribe, payers unwilling to pay (e.g. $300/mo indefinitely), and patients who don't like the non-trivial side effects of existing options. Perhaps future drug discovery will alleviate these issues.
I think this is a fair point, and the final section of the post attempts to address this. For what it's worth, most studies do not selectively draw from a highly-motivated cohort - participants are rarely the most severely obese, and they also usually aren't people who are actively looking for clinical trials or free weight loss programs. Money is also controlled for, since the interventions are almost always offered for free, and any time or work involved with the study will be compensated as well. From my perspective, if anything, the motivational factor may actually be _too_ well-controlled in many of these studies, so the observed result does not translate properly to real-world settings.
Yup - there are several studies [1][2] that have looked into this, and regular weighing (e.g. daily) looks like it has a small but significant effect that helps with weight loss.
Most of the studies I looked at would "enforce" diets through direction, coaching, and meal plans, while a few provided meals to participants. Very few required daily calorie tracking. So while it is possible there is an adherence improvement for certain diets (e.g. low-carb has better satiety, better adherence, and therefore better weight loss), that improvement doesn't appear to be significant enough to regularly appear in research studies. Of course, one could make the methodological critique (e.g. "the studies aren't good enough yet to detect this"), but at least it seems to set a ceiling on the magnitude of a possible adherence improvement.
This is a good point - I'll caveat this in the post. Most medication/placebo studies tend to involve a bit of coaching in addition to the medication, usually 1 session per month or less with instruction on diet and exercise. So the 25% result is essentially "coaching-lite + sugar pill", which may be considerably more expensive than a simple placebo. This result could be much weaker if it's just the placebo alone.
The Science rebuttal raises an interesting philosophical question - if evidence about industry influence is not slam-dunk, as it asserts SRF's influence here was not[1] - then to what extent is it appropriate that meta-conversations about nutrition science invoke external "machinations" to strengthen scientific shifts? For example, if, say >95% of the shift from "bad fat" to "bad sugar" was driven by improved scientific methodology (which the Science article seems to suggest), it seems like a risk of disproportionately highlighting the remaining negligible influence may artificially support the new status quo.
[1] "As we have also shown, the sugar industry approached Hegsted only after learning of
the results of his dairy industry–backed study suggesting that fat and not sugar was a factor in heart disease. “There was no, ‘We’ll get money from them and make the results come out this way,’” recalled Lown, who worked in the department. “It didn’t happen that way,” he said."
Out of curiosity - when you stopped taking the meds, do you feel like there was some residual craving-preventive effect that remained? I'm surprised medications like this aren't prescribed more frequently, and my assumption is that payers / doctors aren't keen on having their patients use it indefinitely (which I'm not sure is necessarily a bad thing, if there are no side effects)
Thanks for clarifying - this is an important point and seems to explain why there was no noticeable weight loss.
From the paper: "During the test phase, participants’ energy intake was adjusted periodically to maintain weight loss within 2 kg of the level achieved before randomization."
Yup - this line of reasoning is also expressed by Dr. Hall, in the slide here [1]. To answer your question, the study specifically states: "On average, body weight changed by less than 1 kg during the test phase, with no significant difference by diet group in either the intention-to-treat (P=0.43) or per protocol (P=0.19) analysis."
It is interesting that the study authors highlight the higher energy expenditure as an advantage for weight loss, but do not seem to directly address that they did not observe this advantage during the 20 week study period.
In case anyone is interested, this thread includes a more specific set of methodological critiques made by Dr. Kevin Hall at this week's ObesityWeek conference (unfortunately does not include Dr. Ludwig's rebuttal):
https://twitter.com/YoniFreedhoff/status/1062760576869371910
Hall's research focus is on mathematical metabolism models, so his critique comes from that angle: it is possible doubly-labeled water measurements of energy expenditure may behave differently in a lower-carb environment, so the observed EE difference may not be meaningful. Ludwig actually briefly addresses this in their published paper, citing a few other studies suggesting that DLW is accurate and carb intake does not mess with isotopic measurements.
Hard to tell if this is a meaningful advance for the low-carb crowd, or just an artifact of insufficiently validated methodology.
Makes sense, thanks for responding. I'm personally partial to chat UI when appropriate, and it does seem like this approach helps reinforce the personalized coaching aspect. Good stuff!
I'm curious about your decision to use a chat interface. Given that the primary purpose is to provide insights and explanation, it seems a traditional UI presentation would be superior -- my sense is that a chat UI is best for two-way communication, as opposed to one party presenting great insights and the other just acknowledging, but presumably you guys have considered other pros/cons.
From an undergraduate standpoint, there certainly isn't a lack of interest in the program - the intro CS course is consistently one of the largest undergraduate courses. Though the smaller department size means certain courses aren't always offered, the quality of instruction tends to make up for this (as well as the possibility of cross-registering at MIT). A side effect of CS being an up-and-coming department is that many CS undergrads come to Harvard intending to study something else (say math, physics, or economics) and thus bring with them their diverse interests and skills to the classroom.
Intent helps people eat better and simplify their lives. Figuring out what to eat is a common struggle of modern life - we solve this. Our app makes personalized meal plans tailored to your goals, preferences, constraints, and routine. We're rapidly growing, people love our app, and we're funded by top investors.
We're looking to hire software engineers to help us craft intuitive user experiences for our native iOS app, and to design and scale our backend services. Creating personalized meal plans is a non-convex optimization problem, so the architecture and technical constraints we use are especially important. Our stack includes Swift, Node, Python, Postgres, and Redis. You'll play an integral role in building and scaling the product, and your compensation will reflect this.
Feel free to reach out directly, will [at] intentapp.com