Jacob 𝞓

25 • Multiple Business Founder • King’s College London • Innovator/Engineer/Strategist

[Jacob 𝞓 @Disruptivism]

SaaS founders - be very careful of both feature creep AND scope creep.

Feature creep makes your product incessantly complex - if you need to, break off a feature into a separate product entirely (Nx monorepo & good DevOps helps with this).

Scope creep will occur in any part of the business or any project - as you do more, you realise there is more to do and the initial momentum is gone. Set a deadline and finish at that date even if imperfect.

Know what your vision is, and stick to it!

Transferable to other businesses too not just SaaS.

[Jacob 𝞓 @Disruptivism]

X algorithm is fixed again

[Jacob 𝞓 @Disruptivism]

@Empress_Brymex It is an intentionally underdetermined phonetic meme, probably designed to generate arguments in the comments.

[Jacob 𝞓 @Disruptivism]

Gathering information:

In business it is wiser to spend 2 weeks gathering information about your company internally and making 1 great decision that benefits everyone in all aspects, instead of trying to guess

Information is the currency of a business

[Jacob 𝞓 @Disruptivism]

When a coke bottle explodes even after it’s accidentally shaken, we still somehow blame ourselves for being careless

Product delivery doesn’t have to be anywhere near perfect

Give customers what they want, at the right price, and they’ll solve how to use it

[Jacob 𝞓 @Disruptivism]

The appearance of choice masks the reality of submission

Most people don't choose freedom, they choose who to follow

[Jacob 𝞓 @Disruptivism]

in nandos rn & they ran out chicken, how

[Jacob 𝞓 @Disruptivism]

@bordzra Yes it is potentially fraudulent misrepresentation, and entering contract under false pretenses, depends how you sell it. Consider an accountant who uses his qualification to pass work to someone who isn’t qualified

[Jacob 𝞓 @Disruptivism]

2 years.

time moves quick, but sometimes u gotta put your head down and do what is required.

no clean plan, no daily routine, just day in, day out, working on breaking the barrier to freedom.

clients, customers, banks, competition, accountants, tech infrastructure , pricing strategies, management, sales, marketing, branding, m&a, negotiation, therapy, tax, reinvesting, cashflow, churn, legal, contracts, employees, planning, strategy, reflecting.

there is no rule book apart from going out there and fighting the things in your way. letting go from the outcome and simply doing what you know is necessary.

[Jacob 𝞓 @Disruptivism]

It isn’t about input it is about your output

[Jacob 𝞓 @Disruptivism]

If u are always the last person to send a text in a convo ur life will accelerate 10x.

[Jacob 𝞓 @Disruptivism]

When starting out it’s best to think of money as a reward for solving problems that other people have

[Jacob 𝞓 @Disruptivism]

Insider tip from research institutions… you don’t need to be INSIDE the university to work with the professors.
Yes it helps because if you are paying the institution they have an obligation to help students

But you can email them and show them skill, and you appear in their best interest (“I’m extending some of your work independently that you authored and I have some promising results so far, would be cool to setup a zoom call to see if we can collaborate?”) - you will be able to find a few contacts in the space FOR FREE.

If you’re really charismatic, eager, show good work results etc. then they’ll be happy to share ideas with you and foster a relationship. (Maybe not ALL senior tenured professors but ones that need more output would happily reply to an email like this). Send 50 emails and I bet you have a good 5-10% close rate.

Random note, I was reading Plato and had a few questions, emailed professors in Philosophy I found online and got loads of replies back and had a great discussion with a few of them about Plato’s work!

Crazy thing is that some PhDs will have less contact with their supervisors.. I’ve heard people only contacting their professors only a handful times of the year!

Professors are so niched and specialised that maybe 10 people in the WORLD know what they’re doing to some understanding/intuition level. So when someone comes along who knows what they’re doing and how to make it better it’s like a rare opportunity for more research.

The only downside is the resources… gonna be hard to get onto the university HPC clusters needed to train ML models.. but if you show good work, even purely theoretical, and it’s in their best interest, I’m sure they would even make an exception or they might even help you apply for a grant.. or you share them the code that they schedule on the HPC for you.. loads of ways to work it out.
The universe will bend to your will.
Remember poor Ramanujan 100s of years ago.. simply sent a letter with a few mathematical notes on and got into Cambridge

Research does not discriminate in STEM (unlike archeology or marine biology). It’s 95% a metricocrasy.

[Jacob 𝞓 @Disruptivism]

@elonmusk What happened 240 years ago

[Jacob 𝞓 @Disruptivism]

Leverage & one to many marketing + not giving a fuck with other ppl think + speaking ur honest thoughts

[Jacob 𝞓 @Disruptivism]

Ppl don’t work for you they work for the business

Mission more important than anything

[Jacob 𝞓 @Disruptivism]

Give this week everything you can

Your future self will be grateful

[Jacob 𝞓 @Disruptivism]

A software startup is an engine that develops a new business - build the engine and with the right mindset you can make anything

Validate early & quickly - a CI/CD pipeline with dev/stage/prod environments

Only develop features that have been requested from a potential customer

When competition copies you, double down because they are also investing

Ensure your product can scale before you get first 10 users

[Jacob 𝞓 @Disruptivism]

SaaS founders.. the business impact of shipping an affiliate feature is much higher than shaving deployment time down further.

[Jacob 𝞓 @Disruptivism]

Multi-model orchestration is emerging as the optimal strategy for developers using LLMs

Single-model approaches inevitably encounter local optima in their latent representations, unable to solve even simple programming problems

Switching between different architectures allows for the widest exploration of distinct manifolds in embedding space

[Jacob 𝞓 @Disruptivism]

Here's the business playbook:

1. Define your "Why" Not what you do. WHY you exist. Write it in one clear sentence. Put it everywhere. It’s branding but just for your staff! You have external branding, you need internal branding too.

2. Turn business values into behaviors "Innovation" is useful but "We challenge assumptions daily" means much more

3. Hiring:
• Skills = teachable
• Values = are they the right person?

4. Trust completely:
• Share a vision
• Define success metrics
• Then get out of the way
• And fire quickly

Real example:
Old: "Work 40 hours a week, you do X, Y, Z"
New: "Our mission is [X]. How would you help? Cool, go do it”

The reason?
When a mission is clear
Methods can be flexible
Results take care of themselves

The best employees don't need management. They need meaning.