I would recommend no, and I've been consistently recommending against AI/ML bootcamps unless you want to transition to a non-SWE AI role. I posted a poll result here a few weeks ago that shows at the top tier tech companies about 90% of engineers weren't evaluating for AI skills in interviews and had no plans to.
MY STRONG ADVICE TO ANYONE WITH EXPERIENCE WHO WANTS TO DO AI: get a normal SWE job at a big tech company that has a lot of AI product or infra and learn internally. They all have thousands of amazing engineers working on AI, breaking confidential research not released to the public or discussed outside, and many have internal courses on AI... it's like better than doing a course in every way.
In the face of tanking enrollment, a number of bootcamps have added AI/ML standalone options for existing engineers: BloomTech and Codesmith being two of them.
They are popping up as a…
I have the perspective of:
1. being a silicon valley outsider who broke in
2. working at Meta from 200 engineers to 10000 and learned a ton about how the sausage is made
3. I started a mentorship program to help engineers prepare for interviews and many people went to bootcamps in the past so I know about them and their pros and cons for your career down the road.
The story of why I'm here.
It all started when a bunch of people applied to my program claiming to have about 6 months to a year of work experience. When I interviewed them their stories all fell apart quickly and I realized these are all Codesmith graduates and the work experience was actually 3 week long group projects and when I confronted someone they said that they were coached into how to talk about it like it was months of work experience.
I did a deep dive and found a Reddit post from 2019 from a tech hiring manager…
If you are cool confidentially DM'ing me your LinkedIn or sending me an anonymized resume on DM I can comment and pattern match what I see right now. It's close to call and depends on your specific experience.
If you have 2 YOE at a good company and ideally for promoted once in that time, then you are in good shape to get a new job.
Now with RTO remote options are less plentiful but they do still exist. The same companies as pre COVID and a few more like Airbnb and Coinbase joined in.
But extremely competitive and you have to be on your top form.
Having the promotion under your belt at big tech will help you hit mid level.
If you don't have any promotions and still entry level, you are going to have a very hard time finding something remote and I would focus on getting promoted first.
If you were hired in as mid-level and had good performance reviews you'll also be fine.
I got feedback that the sub was too doom and gloom and I was trying to post something more optimistic.
You have a lot of experience, so what do you think of the argument?
I've ridden about 100 miles total and it costs about the same as an Uber or Lyft.
I also make notes about all of the things we need non-ML engineers to do every time I ride haha, and I thought the list was so interesting that I wanted to make this post.
The AI future right now: I took a self driving taxi home tonight in San Francisco, like many other nights, and passed by 22 other self driving cars. What this means for YOU is extremely complicated.
This post isn't about bootcamps but rather it's about why technology is so exciting. I can't say a bootcamps are the answer but if you are passionate about technology, I hope this motivates you to keep trying to figure it out.
Ask me questions and share your thoughts!
---------------
I'm lucky that I can zoom around San Francisco in Waymo self driving cars. They sure make the newspaper headlines, but the day-to-day ride is a lot more nuanced than any article or headline would make you believe.
It's not a secret how the underlying machine learning works:
- Add dozens of sensors to a car
- Have humans drive for millions of miles in the cars, record data on those sensors, and send that da…
I think it's really fucked up how someone making an anti establishment argument is manipulating the argument itself... there are are a number of accounts on this thread connected to OP's company.
More of the same in politics.
You said you've been working on AI stuff for 2 years, so why do you understand it better than a fleet of experts who have been doing ML and AI since the early 2000s? Assuming just one of those people has "unfakeable empathy", wouldn't that person be in a place to be a leader on ML or AI?
If someone new wants to get into AI/ML and does your 3 week course to learn the basics of Gen-AI, is empathy enough to supplant someone with the same background but has been doing it for 15 years? Why do assume everyone who isn't in tech has empathy to bring to the table?
Based on that argument you should start a school that doesn't teach any engineering skills and only teaches empathy? And if it's not teachable and something innate, then a school that identifies and nurtures people with innate empathy is great for finding a few leaders of tomorrow, but not accessible to everyone.
This person (who created their account today and uses their full name) works for the OP and reports to him directly at his company - whose homepage claims to create the tech leaders of tomorrow - also relevant, no? Apparently not for either you or the OP in their response.
I think [Klarna](https://www.fastcompany.com/91039401/klarna-ai-virtual-assistant-does-the-work-of-700-humans-after-layoffs) is the canonical example. They fine tuned ChatGPT 3 models on their customer service people and then replaced many with the new model.
I don't know if these were full time employees or if they were located in Europe, or how much they saved by doing it.
Many sides to any story, this isn't my AMA but I have many thoughts on this haha.
If it was in Europe, it's incredibly hard to just lay people off overnight like that haha.
It would be easy to perhaps terminate an external contract you had that had outsourced 300 automatable jobs that were straight up replaced by AI doing the same thing.
I know of two instructors who worked as SWEs and then went back as instructors at Codesmith. I can't speak to their personal situations, but they seem like at least not bad engineers. So I wouldn't say that's universally or unanimously true.
If a decent paying job drops in your lap and you otherwise wouldn't have income, I can see it being an okay option while you job hunt.
Well Launch Academy just gracefully paused indefinitely and another program gracefully closed to preserve its legacy.
So I agree there are other ways.
Two reasons
1. when you pour like 10 years into something, it's just so emotionally. hard to admit what's going on and I can see like not wanting to let it go. I'm not saying that's the best decision but I empathize with that person.
2. Big external companies who are third party CEOs. if I'm a CEO and I get paid a ton of money for keeping something alive for a year I might just want to keep it. or if I'm a big investment firm that owns those company like BlackRock owns Simplilearn, then there are pressures to maximize profits for shareholders as long as you're following the law.
I would suspect this is both app Academy being a really special place back in the day that people really respect and was second that there's a new CEO that'…
I was being sarcastic, but people are paying bootcamps because they think App Academy or Codesmith will get them a $120K job in 12 weeks for $20 to $30K.
So if that's what they are expecting, then what are bootcamps missing - well they aren't getting people jobs quickly anymore... haha.
If people cared about HOW the bootcamp works they wouldn't have signed up in the first place to pay $20K to do a udemy-type course taught in large part by recent graduates.
I really hope the people impacting Whitehouse policy have more experience that a few years as an ML engineer, even if they are an ML engineer from MIT.
I have about a dozen friends who worked at the Whitehouse in past administrations in various capacities in their post-engineering lives and they had years and years of experience in the trenches and tremendous empathy to interact with people who didn't have that experience.
It was tremendously challenging for them, and it takes a really long time to even know if you had impact.
But they everyone there were true experts who were the best in the country at what they do, from every field, coming together to try to solve problems.
A person bringing a unique background to the table AND who has years and years of tech industry experience, would be an asset at the table, but it's far from the only answer to this.
And guess what! Some of th…
So do you think it's likely that a 80% graduation rate is of people who either graduated or dropped out, but that a lot of people might be paused indefinitely and not counting?
That's what I would expect for sure. I work on an interview prep program that is self-paced part-time and people take all kinds of vacations, pauses, ramp up and down, etc... so I would expect that too.
But all the marketing has such high rates that I'm trying to figure out if a piece of the pie is missing.
FWIW, I think their report is far from perfect but the fine print does outline the methodology. They only include "placed graduates who responded to the report" - which is fine to me, but as a consumer you should know that in interpreting the numbers.
A lot of the posts here are from people who are like just finishing their early sprints and oftentimes they're offering discount codes and stuff.
I think you're one of the first people to actually talk about the program after finishing it, so I'm curious if you can comment a bit on how many people made it all the way through and for people who did drop off.
Remote self-paced programs tend to have much lower completion rates as life circumstances come in the way much more than when you carve out 12 weeks of your life to dedicate to a full-time immersive boot camp.
So I'm absolutely not expecting every single person to finish. I'm actually shocked that they publish such high graduation placement rates so want to know more about that.
Time and time again people have tried to lower costs and it results in a worse experience.
I mean while we do interview prep now, the fundamental technology we are building could work for other areas.
But it would be turning the entire industry on its head.
The problem is humans.
It costs too much to have humans in the picture to lower costs.
What do the bootcamps do? Replace them with code and AI that is half baked and not as good but costs 1/100th the cost.
I think GOOD product can do it but. But GOOD product isn't built out of desperation to cut costs when things are bad. This is my concern with these changes that happen post layoff at bootcamps.
I would need about $1 to $4B of funding to make our product able to replace college CS degrees and VCs won't give us that money until we can prove that at a smaller scale.
And this is the conundrum. Building product is an art that cos…
So Leetcode the site came out in like 2012 or something and these companies have been doing these kinds of interviews for years.
It's not actually about the Leetcode, it just feels that way when you look at the interview structure zoomed out like an alien would with no internal context.
Someone told me once that they thought a certain tech leader wasn't very smart (objectively) compared to others. All of the people in power in tech are at least very smart objectively, but each one has a different perspective, experience, and strengths that come out in different ways. Being smart is table stakes, but when you are comparing them all head to head, someone might appear smarter than another.
Being an engineer at a top tech company is similar. Being able to solve problems is table stakes and then what makes you a good engineer is all the rest of the stuff about you.
The way these companies…
We only work with people that have worked in industry for a while, and there are dozens of things those people thought were missing from their bootcamps.
It's a harder question than it sounds because the common thread in the comments I hear is that people had NO idea what they didn't know when they graduated from their bootcamps.
Like "Codemsith told me I did system design after one lecture to be at the mid level and I knew almost nothing at all about it"
If you actually address all of the gaps, you would end up building something that looks like a college degree in CS, maybe some kind of 2 year long college-like program that would have to cost 5X the current bootcamp prices (I.e. what college costs)
Hmm BloomTech (where the CEO of App Academy used to work) also used Canvas. In all fairness she was able to help them survive instead of completely shutting down so if App Academy was headed for shutdown, maybe she's trying to save it like BloomTech.
In my opinion the layoffs in [March 2023](https://www.youtube.com/watch?v=NmIBwP6tBh4) should have been a giant warning sign to run. I know they promised things would improve and get better, but it was prudent to give them a chance to prove that instead of joining anyways and hoping for the best.
Codesmith had layoffs in 2023 and again in February 2024 where it drastically[ cutback programs](https://www.codesmith.io/blog/community-update-doubling-down-on-remote-learning-timeless-pedagogy-frontier-tech). Most of the things promised didn't happen (or didn't happen to the degree stated) and four more employees left in the past few weeks. Yet…
It might be this: [https://www.reddit.com/r/codingbootcamp/comments/1fu01j8/the\_app\_academy\_update/](https://www.reddit.com/r/codingbootcamp/comments/1fu01j8/the_app_academy_update/)
Probably not no. Examples:
1. copy paste a large block of code that doesn't follow conventions of other code
2. copy paste a block of code to a blank pad
3. copy paste large amount of code and change variable names and remove comments (or change comments)
4. copy paste code that has comments referencing where it's from (some sources do this and they paste "From X" at the end, and then the person removes that.
Now all of this sounds silly but in a 90 min session and you do this once by accident and then "undo" to make it go away... won't work, they can see that!
CoderPad can be either full stack or just single-file DSA questions. It has historically been DSA focused.
I would ask the recruiter or the company.
They have a new/beta proctoring feature, but I would guess they aren't using that or logging what you do OUTSIDE of CoderPad. But they can playback keystroke by keystroke what you do in the pad and your presence.
I've watched a lot of recordings and you can kind of tell when someone is cheating or otherwise, long pauses doing nothing at all for example, or copy pasting partial code and working with it like you don't understand etc... You can identify suspicious stuff through copy paste, and clicking, and then have a human look over that section.
I would ask the recruiter for starters though to see what type of assessment it is. Instead of worrying about being tracked, just plan on doing it authentically and you'll be fine.
I can speak for Formation. We are not an alternative to bootcamps at all and not even an option for CS grads who don't have SWE work experience yet. If you have SWE work experience and are considering a bootcamp, then we might possibly be an option, but even then, not a slam dunk and I might still recommend a bootcamp depending on your situation.
We have a very small team and we don't have a lot of online presence, so want don't want people to misunderstand what we do.
For example, a non trivial percent of people come BACK to Formation in the future and pay the full amount to do the same thing AGAIN. This is a good example showing how completely different it is than a bootcamp... it would be like doing the full Codesmith Immersive for $22,500 twice two years apart... makes no sense.
Second, we don't teach anything or have any classes. This is SUPER important because people should not…
There isn't a home for those (add Pathrise, Interview Kickstart and Outco to the list).
I'm the co-founder of Formation, and we call ourselves 'interciew prep and mentorship' rather than a career accelerator.
So the discussion here tends to be from bootcamp grads later on in their careers.
The discussion in CS majors is mostly about Headstarter and Coachable because this appeal more to that demographic.
Pathrise is big with people on Visas.
Interview Kickstart in is India and bigger in those areas.
Interviewing.io and Hello Interview are more a la carte Interview Prep and more in the Leetcode sub. Formation is also mentioned in the Leetcode sub sometimes.
Very bluntly, there are all entirely different programs and audiences with a lot of overlap between various ones, but no single gome really makes sense. People tend to be looking at 2 or 3 of these and which 2 or 3 are differen…
How does this compare to Codesmith and Hack Reactor? 20% in six months doesn't seem that terrible for 2023 compared to the other good ones. It's definitely much lower than historical numbers so everyone should be transparent about these because people shouldn't be mislead about placements for sure.
1. sometimes less is more. having more people around can cause more communication challenges, require more guardrails in the code etc. as not necessarily like an overall unit of output.
2. the impact that an engineer can have in general isn't just like 1x 2x it can't be like tedx or more of business impacts that someone can have. so some of those Engineers are just not dividable into any number of new grads to replace them.
3. new grads aren't expected to have positive impact for the first 6 months to a year as they ramp up. so if you have one person who's having tremendous impact, it's not really an option to just replace them with five new grads because those five new grads are going to need a while to ramp up and they'll be actually not just the loss of the more senior person, but even it more additional loss from the ramp up.
4. in the boom times companies had so much momentum and…
Next time you buy a Lamborghini for 1/10th the cost and find out that it's really a cheap car with a sticker covering up the logo let me know.
And then when you find out that the company sells all the premium brands and all are the exact same cheap car with stickers covering up the logos.
And it worked because the people buying them have never seen a Lamborghini before and trusted the salesman because they bought it at this building adjacent to the dealer.
The alternative is you could just buy a reliable and highly rated cheap car that gets you from A to B instead...
There's a fundamentally broken aspect of the entire education industry - both colleges and bootcamps.
You aren't paying for quality of education, you are paying for a name and an entrance bar. It just so happens that if you have a strong brand and the highest entrance bar that you end up with the most brilliant people all networking with each other and who go on to great things. It also means you have the best teachers who want to work at the best school and get the easier research funding.
Stanford is an example of this working.
Bootcamps are examples of where this entirely falls apart.
People should be paying to learn - but they aren't, they are paying for a label and community. If the thing doesn't deliver Stanford-like outcomes, the entire thing falls apart.
Codesmith is the closest example of this. It was trying to be the Ivy League bootcamp and unofficially had that reputation…
I definitely agree. product market fit is number one. That's table stakes. Companies can succeed with product market fit and lack of talent if they're able to catch up because their product market fit is so strong.
If you have super strong product market fit, like some boot camps did during the boom, then you shouldn't be hiring back your students. you should be hiring Meta Engineers for $500,000 a year. And if you can't afford that then you don't have product market fit yet.
If you have all the talent then you have one of the boxes checked but you still need product market fit. but this is why former employees or these companies get so much funding because they've already checked off some of the boxes. and the biggest question is just product market fit where people who don't have those backgrounds have more unknowns for these check boxes.
I completely agree that it's not even clear…
One of the challenges in the EdTech industry is there is a frenzy to incorporate AI and all the investors are about AI... but the EdTech industry doesn't attract the best of the best product people and engineers the way that OpenAI, Google, Meta hire in droves and pay $1M a year.
Sure there are good engineers in EdTech, high potential engineers maybe too, but not the best of the best who have the experience of building and scaling products.
Why?
There are no/few EdTech companies that have gone past $10B and a lot of those larger don't have the growth rate or margins to attract those top talent and pay them competitively.
So the best people generally don't go to these companies.
But the market is one of the largest in the world, trillions so people keep trying and keep getting funding.
Without trying we're not going to make progress... it's just hard to experiment with people paying…
Speaking from Formation, we've been seeing increasing interview activity since July but we're calling this the 'new normal'. We expect to see more layoffs and more hiring in Q1 2024 and uncapped hiring surges like in 2021-2022 mid COVID... efficiency is the new normal.
Also no one should say anything until after the election settles. Things could change really quickly!