I’m semi-normie (MechEng with a bit of Matlab now working as a ceo).
I spend most of my day in Claude code but outputs are word docs, presentations, excel sheets, research etc.
I recently got it to plan a social media campaign and produce a ppt with key messaging and content calendar for the next year, then draft posts in Figma for the first 5 weeks of the campaign and then used a social media aggregator api to download images and schedule in posts.
In two hours I had a decent social media campaign planned and scheduled, something that would have taken 3-4 weeks if I had done it myself by hand.
I’ve vibe coded an interface to run multiple agents at once that have full access via apis and MCPs.
With a daily cron job it goes through my emails and meeting notes, finds tasks, plans execution, executes and then send me a message with a summary of what it has done.
Most knowledge work output is delivered as code (e.g. xml in word docs) so it shouldn’t be that that surprising that it can do all this!
If, during a state of defence, the need for civilian services in the civilian health system or in stationary military hospitals cannot be met on a voluntary basis, women between the age of eighteen and fifty-five may be called upon to render such services by or pursuant to a law. Under no circumstances may they be required to render service involving the use of arms.
“Human beings, who are almost unique in having the ability to learn from the experience of others, are also remarkable for their apparent disinclination to do so.”
Obviously, this is not going to make up for the loss of the broader ship building industry, but it does show that the UK is thinking about maritime technology as a key strategic area.
That’s the premium solution that some, but not most, can afford and only makes sense if the cost of repair is less than the cost a new unit. Travel just makes that tip in favour of getting a new system with a $10 deposit.
Most of the people we’re talking about here are subsistence farmers who pick up casual labour at a local farm. Income is sporadic and seasonal.
That was the initial brilliance of the PAYGO system, it allows users to pay off their device sporadically I.e. they buy units when they get paid and that goes towards paying off an asset that in theory will then provide energy at 0 marginal cost. Turns out that last bit isn’t true.
Here the VC story is important, these companies were meant to be high growth and giving significant returns. We all know how that ends.
> You would be surprised at the amount of product repairs that are deemed not worth solving in a developed country that you can sort out in a couple of hours in a developing country.
I have been in the past, but not anymore. No one is saying people aren’t resourceful but there is a significant barrier to entry when it comes to electronics repairs for the general population. One part of what we provide is an off-grid repair lab bundled with our new education offering so it’s very much knowledge + tools.
Cheap crap = lasts only as long as the warranty period.
I did a survey in partnership with a the African Leadership University in a Rwanda, where we surveyed people living in two rural villages and found 90% or units had broken within 3 years of purchase. This is the logical end point when 1/5 stop working after 6 months, which you can find in Cross and Murry 2018, linked in other comments.
10x mark up (i.e. the mark up on cost of the unit) comes from knowing that the COGS for one of these units is ~$20-30 and the premium sellers sell up to $300.
Sure it’s at the top end of the range but 10x markup on each unit is not an exaggeration, let alone a massive one.
Gross margins are indeed tight but that’s is a separate issue to markup. You can sell at a huge markup and still make a loss: for example if the default rate of loans you make turns out to be much higher than you expected.
Yeah not much has changed, it’s simple technology that was developed in 2005. Main innovations of the PAYGO model in recent times was the ability to use the payment data you collected to offer other types of loans. One of the previous biggest companies in the space now sells mobile phones using PAYGO.
Technically battery chemistry has obviously moved on but we are talking a device capacity similar to a medium power bank. How much innovation have you noticed in power banks recently?
Panels are big problem from a e-waste perspective as they very difficult to repair.
Batteries failures are repairable. Usually battery packs will be 2+ LFP 18650s or 32700s. If one cell goes bad the the whole pack goes but the others may be fine. Just need to test and match cells and you can make new packs.
I can’t remember exact recovery rate for cells, I think it is something like 40-60%.
Dealing with these batteries at end of life is a challenge, but that’s a global problem.
Still a lot of legacy Sealed Lead Acid batteries around but these are very recyclable.
For sure recycling is improving. That’s different from repair though.
The current cycle is
1. sell product
2. wait three years for it to break
3. Go back to 1.
The impact of the recycling can lessen the impact of that but it definitely doesn’t eliminate it. That’s just on environmental scale, think about the financial impact of carrying this debt for years on people earning $2 a day.
Also important to note that a lot of this is contingent of legislation that implements things like Extended Producer Responsibly (EPR) where you essentially have an additional tax on producers that gets used to fund collection. Kenya implemented this for the first time 12 months ago [1], so we will see the impact over the next couple years.
Re solar punk, my personal vision is that you basically teach people how to build and maintain these systems themselves by running solar tech bootcamp and giving them off-grid tools.
They then have tools and skills to fix anything without the need for the grid. Train 100k people and have them maintain these systems using a decentralised approach.
In fact, as part of our training we now have e-cooking stove suppliers who deliver training on their stoves to our students.
The economic impact of this cannot be over stated.
1. You are giving people the ability to 4x their income as repairers
2. You are saving the people who are getting new systems, instead of repairing them, multiples of their yearly income.
I didn’t say, nor do I assume, people are inept at repairs. I said that the communities that are being targeted with off-grid solar don’t currently have the skills or infrastructure required to maintain these systems.
It’s something I’ve seen with my own eyes and that I’ve read in academic literature as a widespread problem. Cross and Murray 2018 [1] being one the first papers to talk about it, I saw it myself for the first time around that time in Tanzania.
I stayed in a village where each house had at least 2 broken solar lanterns stored in a corner (like those old routers people love to keep).
The next closest repair shop was first 30min motorbike ride then a 2 hour bus ride away.
This was a village of 8,000 people.
Yes, the person with a diploma from the local technical college can fix a lot of things but they live in the local town with grid electric etc. They don’t live in these remote rural regions where off grid is so important/impactful.
The solution is to teach more people how to design, build and maintain solar energy systems so that the skills are embedded in off grid communities and give them the tools to carry out the work. You can do a lot with a soldering iron and good grasp of electronics!
Those who are talking about market opportunities, yes it’s big!
Bottom up calculation: average $10/repair x 250 million potential repairs = $2.5B market.
Problem is labour shortages and supply chain, as stated in the report. Both hard problems to solve.
We’ve been working on getting the labour shortages fixed and I personally believe that you can also skip some of the supply chain problems by localising labour.
For example: when we train people they can 4x their _household_ income within 6 months. This is young people who didn’t have an income before and are suddenly earning 3x as much as both their parents combined.
People just don’t know how to fix these things and when someone finally learns how, they can absolutely rake it in.
It’s actually insane to me how much education can be such a massive multiplier in this context!
The repair service is really expensive for subtle reasons.
First of all there is the repair itself, but there isn’t any collection service so you need to travel (often a full day) to get to a repair centre. Then travel back. Not open on the weekend so you have to do it during the week, meaning you are also loosing 2 days of income. Then you have to go and get it once it’s done.
Total cost of repair for people using these devices might 10+ days worth of income if you include the opportunity cost.
That’s why we are training people to fix these systems within their communities.
Regarding parts you can get second life batteries in Kenya for $1-2 per cell from people like Acele Africa[1], so you can get total repair cost down to ~$10 (that’s ~3% of original purchase price)
People are paying off these devices and then once they have paid them off, they break and people in these areas don’t have the skills or resources to fix them.
This has led to over 250 million of the units lying around broken in peoples homes, leading to solar being one of the fastest growing e-waste streams in the world.
It’s hardly solar punk to sell people cheap crap at a 10x mark up that pretty much immediately breaks once the warranty period is over.
I’m semi-normie (MechEng with a bit of Matlab now working as a ceo).
I spend most of my day in Claude code but outputs are word docs, presentations, excel sheets, research etc.
I recently got it to plan a social media campaign and produce a ppt with key messaging and content calendar for the next year, then draft posts in Figma for the first 5 weeks of the campaign and then used a social media aggregator api to download images and schedule in posts.
In two hours I had a decent social media campaign planned and scheduled, something that would have taken 3-4 weeks if I had done it myself by hand.
I’ve vibe coded an interface to run multiple agents at once that have full access via apis and MCPs.
With a daily cron job it goes through my emails and meeting notes, finds tasks, plans execution, executes and then send me a message with a summary of what it has done.
Most knowledge work output is delivered as code (e.g. xml in word docs) so it shouldn’t be that that surprising that it can do all this!