No programming required, just plug and play. Is that wishful thinking for nonprofits or the only way they will be able to take ownership of AI processes in the future?
Will social sector organizations be able to harness the potential of AI and machine learning to help solve the defining issues of our time with localized solutions or will they continue to look at expensive third party vendors or consultants, looking towards innovative foundations, donors and investors to help fund it all?
Look, it’s not too wild a notion to think that no code solutions (or low code options) will be readily available and more affordable in the coming years, especially if we realize that many vehicles currently exist if we are willing to look for them, adopt them and ultimately utilize them to their utmost capabilities.
It’s not too wild a notion as ‘no code’ options are readily utilized by us for a variety of our work already. White label companies are out there for you to build your own apps and crowdfunding platforms too. Simply building out and personalizing on top of their templates with a simple drag and drop and all of a sudden you look like one of the top 10% of cutting edge nonprofits in the country.
For $50 a month you can take donations, set-up chat bots, send push notifications and other custom content to engage your clients, donors and volunteers.
Other traditional nonprofit disciplines will soon be turned on their head soon too including:
Fundraising – companies such as piLYTIX, DonorSearch Aristotle and the Futurus Group are already building out algorithmic modelling that is helping organizations identify, build and review donor pipelines.
Operations – Multiple companies are already out there supporting enhanced business operations and workflows.
Many of these groups will also begin to identify more as AI companies rather than a software company. All you have to do is look at the jobs at Canva currently being advertised, machine learning engineers, data analysts and computer scientists.
But your quintessential nonprofit isn’t going to have data scientists and the like on staff, heck 40% of for profit companies (according to a Deloitte survey), state AI technologies and expertise are too expensive. So how can AI be democratized to be widely available at a low cost?
The answer lies in no or low code AI.
No Code AI – platforms with visual, code-free and in most cases a drag-and-drop interface to deploy AI and machine learning models. This option supports non-technical users to quickly classify and analyze data to build effective prediction making models.
Low Code AI – perfect for IT folks on staff with development skills. These platforms require technical knowledge, but allow them to work faster (like an IKEA for building algorithms).
The inherent benefits of these options include:
Usability: Technically (excuse the pun), it allows for anyone in your organization to find an AI solution. No code platforms are designed with non-developers/non tech savvy folks in mind.
Speed: Building off templates with drag and drop options allows for rapid ideation and real time changes (based on both the data, UX/UI issues and of course, typos).
Risk: No-code tools are built on top of proven scaffolding thus mitigating risk of building something from scratch and having it not work.
Scalability: AI gets better with more data. Fuel the machine and don’t pay a dime extra for each piece of meta data uploaded or how many users upload said data.
Cost: It’s a much lower investment than bringing on new staff or consultants.
Potential: Template centered platforms can justify the larger adoption and use of AI in the future.
The social sector is probably more leaning on no code ai offerings especially given some key stats that loom large over the sector including:
Awareness & Trust (stats here via ‘The State of AI in the Nonprofit Sector was conducted by PwrdBy’)
Nonprofit practitioners are aware of AI, but have reservations:
- 59% of nonprofits hear about AI from their CRM provider. (nooooooo!)
- 83% of nonprofits believe an ethical framework needs to be defined before full adoption of AI in the sector
They also found that the decision-making process for adopting AI is still very mixed. Most
people believe they need more time with AI before feeling comfortable (63% of respondents)
and 83% of people believe an ethical framework should be put in place before full AI adoption
occurs. 52% of respondents also reported being scared of AI advancements.
In addition to how they hear about AI, this largely falls at the feet of organizational technology champions (55% of respondents), their boss (47%), and their board (44%).
Internal Infrastructure (stats here are derived from both the 2020 State of Philanthropy Tech from the Technology Association of Grantmakers and the Salesforce Nonprofit Trends Report)
Lack of IT staff is hindering understanding, adoption and opportunities:
- 93% of respondents state a lack of IT or technical staff is a challenge to their organizations adaptation of new technologies. (remember that 55% of the use of technology is actually championed by IT).
- 17:1 continues to be the average ratio of staff to IT staff
- 40% of IT departments do not have any DEI programs (compounding issues of trust)
- 51% of respondents expected to see their IT budget increase in 2021 (and that’s with COVID exposing the limited effectiveness of our service delivery models).
Couple this with a demand shortfall of around 1 million computer scientists in the US alone, we see that engineering talent is scarce, so no code options might be the best chance of onboarding this generational tech to the social sector en masse. It’s just whether the industry has enough verve to see their calls for AI democratization through.
Ok, that’s a lot of stats, we get it. But what are the substantive benefits? Well before we get to those we should put this all in context with some actual examples otherwise this just becomes a ton of unrelatable jargon to those that just want to take advantage of ai to improve inputs, outputs and solutions where the cost and accessibility barriers are stripped away. How do nonprofits and small to mid-sized foundations create A.I. systems using simple visual interfaces or drag-and-drop menus?
Well, much of it’s early uses will be made for binary/rudimentary uses that tackle unstructured data such as:
HR: Analyzing all new employees onboarding documents in real-time and being notified when files are complete.
Grantmaking: Find out if nonprofits actually qualify for funding based on your criteria by analyzing public data (possibly identifying other organizations in the meantime and inviting them to apply).
Grant Reports: If you’re going to make nonprofits write lengthy reports (something tech solutions could end up eliminating altogether), then run the reports through a model segmenting the data and uncover underlying patterns.
A good visualization here from the for profit sector can be seen on the Levity site here.
AI ultimately is something that works in the background and is dynamic, not static, running its models daily and feeding off the inputs of new data from wherever it pulls it from (mainly your CRM – hence why they want more data, and for it to be better categorized and labelled, and why they are pushing new advances!)
As you know AI stands for Artificial Intelligence, but we really need to be seeing it as automation if we are to have our sector lean into it and not be – as the above report stated – scared of it. So if we are scared of automation how are we going to fully grasp or understand the power of AI moving forward?
Firstly there is ultimately potential in power, it’s just how we conceptualize it. Do we want systems that automate the most mundane administrative tasks to allow staff space to think, learn, ideate and grow (not backfilling that time with additional rudimentary tasks)? Do we want to run models that review our donor prospect pipelines seeing if they are ready for an ask? Do we want independent review of our data to identify trends and patterns that we might not necessarily see, especially if it helps us identify clients in need of urgent help or throws up ideas that might actually recommend a new approach to solving some of the most critical social issues of our time.
And that’s just on the nonprofit side.
For organized philanthropy, Imagine purchasing a number of data sets that help inform grant cycles rather than create them based on potentially biased assumptions to trending issues? Imagine modeling that identifies successes (or issues) with a funded project and triggers additional funding to ensure success? Imagine models that review projects and build out their own independent reviews that is then distributed to similar organizations with the same NTEE codes to effectively help everyone ‘level up’.
Now imagine you can build these yourselves. And that tech for good advocates are building them and making them available to the sector at a reasonable cost. While it would still need a broader sector dialogue, it’s a conversation worth having. The conversations around ethics in ai is already occurring across all industries and largely driven by some of our leading higher education institutions, but we need to talk more concurrently about the possibilities of ai for the social sector and how as many people as possible can benefit from it before the power we mentioned earlier becomes entrenched once more at the top and that potential is another opportunity missed to level the playing field and advance society.
At the end of the day, no-code AI is still an emerging market. It hasn’t even begun to turn it’s eyes towards the social sector (beyond government) yet, nor has it even trended towards use case management options (classification problems, CRM, web-builders, business apps) as it focuses on building out its core technologies in vision & speech recognition etc. But it will be coming and nonprofits will be seen as a potential revenue stream.
The point being made here is that we should be thinking about it, thinking about its applications and reaching out to those building out the technology – to partner, inform and volunteer to beta test it’s applications. Companies should also be proactive in reaching out to nonprofits and philanthropy too to help understand how to make their models and templates more dynamic, robust and effective too. For far too long it has been a top down approach with tech companies ‘bringing our sector the solution’ and making us pay a premium for it, when all it has done is simply rebadge their for-profit enterprise model. Let’s get the tools into the hands of those that can make a difference and let us all reap the rewards, but with the novel idea of a shared focus on lifting up those most vulnerable in our society.