Saying “ChatGPT told me …” is a fast track to getting your input dismissed on our team. That phrasing shifts accountability from you to the AI. If we really wanted advice straight from the model, we wouldn’t need a human in the loop - we’d ask it ourselves.
We use AI heavily in our product and development flow. Sometimes we'd encounter a problem none of us can figure out at the moment so some of us would use ChatGPT to brainstorm some solutions. We'd present the solutions, poke holes into them, and go forward from there. Sometimes we don't use the actual ideas from GPT but ideas that were inspired by the generated ideas.
The intent isn't to shift accountability, it's to brainstorm. A shitty idea gets shot down quickly, whereas a good idea gets implemented.
First of all: great that you’re working on bringing innovation to general aviation! During training I also wondered about some technical aspects, as a lot of the tech in an airplane is from almost 60-70 years ago.
However, as a PPL holder myself I think that understanding weight and balance, aerodynamics, flight planning, ATC, airlaw, principles of flight etc. is paramount to keep yourself and others safe.
In the air you cannot simply drive to the side of the road and have a look at what’s going on. As a pilot you must be able to make decisions on the spot. To be able to make those properly, you have to understand what’s going on from first principles. In aviation safety is the primary goal. Statistically more than 80% of aviation accidents can be traced back to human error. Anything that facilitates bringing people in the pilot seat suggesting they need less training is something i strictly oppose.
Despite, anything that helps to simplify aircraft operations may decrease cognitive load and is therefore welcome.
I guess you're aware of the ICAO/FAA hoops you have to jump through to get anything of what you're developing beyond experimental certificate but having seen how long it took e.g. Volocopter to gain their (still incomplete) type certificate it is a long road.
If you'd be able to simplify avionics and create an affordable standardized fly-by-wire kit that alone would be a great innovation for GA especially in the ultralight and light sport aircraft category.
> anything that helps to simplify aircraft operations may decrease cognitive load and is therefore welcome.
exactly. decreasing cognitive load is very important and I think we as existing pilots tend to forget how much of their cognitive load goes to just aviating. But the students who drop out certainly don't.
Just an anecdote (no judgement here; VC has its place): An acquaintance of mine who works for a VC firm once said "Ultimately, VC money is a loan for people who are not bankable".
That really resonated with me as with that perspective I understood why behavior & practices are closer to what you'd experience if you personally need to take out a loan outside of the regulated banking system
The whole "hate VCs" thing is kind of silly in my opinion. VC wouldn't exist if founders didn't want and need capital and have no other way of getting it.
Also, for every other "evil VC" story, there are other stories where founders are really happy with their VC board members, have a strong and positive relationship with their VC partners, and end up getting some kind of positive exit which wouldn't have happened at all without the investment.
Sure, some VC companies may be shitty, some others may be amazing, but this is basically like everything in life. Some schools are shitty, some cops are shitty, some cars, tech companies, managers etc, are shitty. But some of all these things can also be great and awesome too. Thats life.
A - Just as with customer reviews on amazon unhappy customers are often times the majority to leave a review whereas fewer happy customers voice their opinion in the form of a review. I suspect the same goes for VC interactions.
B - Due to the way the VC business is structured the variance in "quality" of VCs is heavily skewed and not normally distributed, tricking our perception of what to expect. In other words i suspect that you have a much higher likelihood in to interact with a very "low quality" VC or absolute "top VC" than with an "average" VC. If you amplify this with A you may get an even worse public opinion.
Nevertheless I think on an individual basis you're always better off if you don't need VC for your business - if you have that option.
VC wouldn't exist if founders didn't want and need capital and have no other way of getting it.
This is too reductive. A lot of founders have to raise because they're competing with other companies with VC funding, often dumping their product on the market at a loss to starve out bootstrapped competitors and lock in customers.
Yeah but imagine trying to bootstrap a product which relies on network effects with no investment and where profitability is potentially years out. Social Networks are a good example here, or something like Uber.
There are many valuable and amazing businesses where self-funding/bootstrapping doesn't work. Or businesses which start out as one thing, get VC money and experiment, try to find product/market fit over a year or two, and then hit it big and create something really valuable to the market.
I agree that many companies would be better bootstrapping and creating a sustainable, profitable businesses, but the truth is, the VC model and ecosystem does enable certain businesses to exist which wouldn't exist otherwise as people wouldn't work for that long for free or almost no income.
Sure I think VC has value in some cases. I think the zero-interest period distorted their role and pushed them into areas that would have better been served by bootstrapped and profitable small companies. I'm hopeful that with interest rates back in a more reasonable range we'll get a better equilibrium.
Cost/price is another key aspect that is not prominently present in the article but is a key decision driver. The assumption on the employer side is that younger folks will work for lower salaries AND can be developed to the companies needs whereas someone with 10-20 years of experience is automatically “more expensive” and has to adapt.
Now there is a catch: even IF an experienced candidate would work for a lower salary, e.g. if they don’t have exactly the matching skillset and need time to learn, the employer may see this as an indication/admission of a lack of competency and decide against that candidate.
It’s a tough challenge and often a lot of irrationality seems to be involved.
After decades of picking up new technical skills as needed I have confidence in my ability to learn new things so I don't waste my time learning any technology for which I don't have an immediate need. Most employers, not all, seem to primarily focus their hiring decisions on 'what do you know' versus 'what have you done'. Naturally, that negates the primary advantage that more experienced workers will have.
Can confirm that you may run into trouble e.g. when requiring screenshot capabilities (via scrot in our case) hence we had to set the WaylandEnable to false as well.
I leave the house for a walk or a bike ride for at least 15-20 minutes which is basically a "fake commute". When i come back it is like coming back home from work and i can easily switch off.
Whenever i don't, I have troubles to switch off from work.
I like that this article is very clear, visual and to the point. From my experience those exact same approaches/models are very useful throughout many situations in business, not only for designers. Consider e.g. sales where you also need to
- show (the customer) how to solve the problem (he/or she has)
- create a clear decision framework (and get the customer on board)
- craft a narrative around our solution
to succeed. There you also need your problem solving skills, a model for decision making and a way to communicate in a way that reaches and convinces your audience.
Saying “ChatGPT told me …” is a fast track to getting your input dismissed on our team. That phrasing shifts accountability from you to the AI. If we really wanted advice straight from the model, we wouldn’t need a human in the loop - we’d ask it ourselves.