The 2025 Bum Thread
- Thread starter mid_gen
- Start date
Basic usage from AWS bedrock or Google collab etc are cheap. It only gets expensive when you deploy a product at scale and go through billions of tokens an hour .
But playing with prompts is about 0.000000000001% of the work of developing a production ready genAI solution
True.
I use Google Collab's Python and Jupyter Notebook, you can load data into it (obviously don't use PII).
Having delivered AWS Bedrock as part of the cloud platform, and had all the fun (eventually risk drove to a private model), I'd look at spreading the experience (and make sure you have your spend limits setup for cloud, just in case).
Was thinking last night - what options I have, the main areas of investment are cybersecurity and AI.
1. Job position at place, non-exec/senior role -> under 50 and demand for demonstrated enterprise experience of AI is only way to get interviews, job market swamped with seekers.
2. Job position at place, exec/senior role -> 50 still a problem, demonstrated experience in AI seems only way to get interviews again, job market swamped with seekers.
3. Go contracting -> over 50 less of a problem, more of a problem is demonstrated experience in AI, job market swamped with contractors.
4. Create own company -> age irrelevant, no different now (so no fear of losing job), find market, start building out (in addition of above), danger is lots of people creating AI companies with lots of noise but little action.
Unless you're looking at research based companies, there's very little investment other than invest a little to bring down costs a lot.
I'm still doing 1 & 2, also adding 3 in the mix and from last night .. 4 seems to be something attractive in building from contract into selling product offerings. The danger is the drowning noise.
In fact it's quite funny.. it's said that older people avoid risk. No they don't, they take risk in a sensible way vs the younger person that are oblivious to the risk.
1. Job position at place, non-exec/senior role -> under 50 and demand for demonstrated enterprise experience of AI is only way to get interviews, job market swamped with seekers.
2. Job position at place, exec/senior role -> 50 still a problem, demonstrated experience in AI seems only way to get interviews again, job market swamped with seekers.
3. Go contracting -> over 50 less of a problem, more of a problem is demonstrated experience in AI, job market swamped with contractors.
4. Create own company -> age irrelevant, no different now (so no fear of losing job), find market, start building out (in addition of above), danger is lots of people creating AI companies with lots of noise but little action.
Unless you're looking at research based companies, there's very little investment other than invest a little to bring down costs a lot.
I'm still doing 1 & 2, also adding 3 in the mix and from last night .. 4 seems to be something attractive in building from contract into selling product offerings. The danger is the drowning noise.
In fact it's quite funny.. it's said that older people avoid risk. No they don't, they take risk in a sensible way vs the younger person that are oblivious to the risk.
Last edited:
Seems to be plenty of investment around, no problems with the startup I'm at and the various other startups we're involved with. Still loads of tax incentives around making investment very low risk for people. You do have to tack AI onto your offering somewhere to pique people's interest but it doesn't have to be the core of your offering.
I wouldn't be trying to get jobs developing with AI. It's so new that as an experienced 50+ professional you are never going to compete with 20 years olds that have been using it just as long.
Besides that, creating something with AI is a piece of ****, I've implemented the AI assistant on our product. If you really want to work in AI then just buy a domain and build something, and demonstrate that you can actually make stuff.
Cool that there's uptick in funding (last year friends couldn't get movement on that and they have built companies). Yep - that same boat in terms of experience is what's the kicker.
I do think, that the dust is settling in AI.. people have the paper cuts and will consider it but it's not the silver bullet (which everyone understood).
I'll finish the cyber course, and have a think (whilst applying for roles).
Last edited:
Problem you're going to have is that having a lengthy CV in one area, and a course in another, isn't valuable to employers. They can just get a fresh graduate who's done the same course, and pay them less and develop them into what they need.
People care about what you've done, not what courses you've completed. I could spend months doing training courses, but what would make me employable as an AI developer is the fact I spent a week building an AI assistant for an application.
Agreed. Yet, it's a self reinforcing circle, unless you build something they see and get excited about.
20 somethings probably don't sit and complain about entitlement due to experience too and just get on with it
Last edited:
Mrs asked have I heard back from applications I've sent - I felt like manically laughing and joking that she's being a little naive - but she's under enough stress.
It would be great to have at least some demonstration of interest, something that feeds back to say that's where you're going wrong. Nope.
It would be great to have at least some demonstration of interest, something that feeds back to say that's where you're going wrong. Nope.
Soldato
- Joined
- 24 Sep 2007
- Posts
- 6,044
This is the way forward if you are quite old (like me). If you have the skills, put them to use for your own profit, not for other people.
For your own business, you should try a few different things simultaneously. One of these should be something that purely has the ability to make money, i.e. nothing fancy, not necessarily AI related, it just can make money. What I am saying is do something simple that can make money (easier said than done). Then if you like the AI stuff, do that as well, but as an additional potential revenue source, because AI is hard, and highly technical things are not easy to make money with if you are one person.
Last edited:
I have a feeling that this exec role is a principal role reporting to the board which could be anything from cutting new ground to troubleshoot. Even if it's a 6-12 month contract it'll be good to get the brain operating at full capacity.
Last edited:
Been approached for a contract role.. turns out it's for the same company I used to work for, reporting into the CISO I used to have calls with.
The role - is bang on what I've done before at a large company and at a senior level sorting out their **** for the CEO. However there used to be fireworks in term of politics between group (who I worked for) and this sub-company. Politics not my doing (between boards themselves), but there's likely some good and bad in the feelings that created. The CISO is extremely demanding, stressed but does warm to you.
The role - is bang on what I've done before at a large company and at a senior level sorting out their **** for the CEO. However there used to be fireworks in term of politics between group (who I worked for) and this sub-company. Politics not my doing (between boards themselves), but there's likely some good and bad in the feelings that created. The CISO is extremely demanding, stressed but does warm to you.
Nothing back from that application.. so carryon up the recruitment..
Finally got a response from the recruiter yesterday after chasing it - the company wanted more legacy experience. Ok, so be it.
Last edited: