When I first saw Dr. Tooba Durazze Ph.D in a video online I immediately became a fanboy. Here was a young woman, with a bit of a hipster aura, and a PhD. from MIT, talking about how neurosymbolic AI is a better model to base your business on than large language models (LLMs).
Dr. Tooba is Founder and CEO of Amoeb.ai. Amoeba uses "advanced AI—deep learning, cognitive AI, and neural networks—to deliver tailored, precise insights that empower smarter decisions," automating "the entire process, making AI-driven decisions accessible without specialized skills or large data teams." It does this by basically installing an equivalent of a Data Scientist in your organization to make sense of, leverage, and ultimately make actionable enterprise data.
I may be swimming out of my depth here (check that, I am swimming out of my depth here), but as I understand it, the difference between LLMs and neurosymbolic AI is that LLMs being prediction engines cannot ascertain causality, only probabilistic correlations. As such, they look at past data to make a prediction about future trends, but they cannot be certain, and of course, LLMs seem to have more hallucinations than a Timothy Leary party.
Thus, while LLM outcomes may sound confident, the fact that the reasoning is a black box, makes it not 100% reliable to run your business on it.
Instead,
Amoeba takes a fundamentally different approach. Instead of relying solely on probabilistic predictions, we fuse two complementary systems:
Neural Networks for pattern recognition across messy, fragmented data; and Symbolic reasoning for logic, rules, and causal structure.
This neurosymbolic approach allows Amoeba to move beyond surface-level patterns into true causal intelligence. The system doesn't just tell you what happened - it explains why it happened and what to do next.
BEYOND LLMS: WHY NEUROSYMBOLIC AI IS THE NEXT STEP FOR BUSINESS INTELLIGENCE, on the Amoeba AI Blog.
Incidentally, this is the same argument Gary Marcus, Professor (Emeritus) at NYU, makes in his critique of the currently fashionable Scaling Model of AI followed by companies like OpenAI and others.
But enough on that. I was lucky enough to meet Dr. Tooba at a MeetUp hosted by Arjun Saxena, Founder and CEO of Humanic, and the host of the AI x Marketing Summit (November 5/6, San Francisco),and had an interesting discussion about her ideas, Amoeba AI, and the future of AI. The video is available on YouTube
Near the end of our interview I asked her about what advice she would give to job seekers in this market, especially mid to late career job seekers needing to re-tool to stay competitive and thrive in the Age of AI?
She had three recommendations:
(1) You should be a "friend of learning" meaning be curious, and not be afraid to be continuously learning. With new stuff coming out every day, don't be afraid to jump in and play. Learn new things.
(2) Have humility in knowing that you don't know everything, because the moment you say you know everything is the moment you fall flat. This is especially important for older workers, who may come across as "know it alls" not necessarily because of arrogance (although there is a lot of that), but maybe due to insecurity, not wanting to show vulnerability in not being up to date. We have to get over that, and be open, collaborative, and not be afraid to say you don't know.
(3) For those who are starting companies (and this could be professionals who start a consultancy, or become a contractor, or start something completely different than the area they worked in, is to create your own blueprint. Learn from others, but don't follow them blindly. Create your own path.
The entire segment of that part of the interview is reproduced below, and the video can be seen on LinkedIN.
Fal: Hi, who are you? What's your name? What do you do?
Tooba: Hi, my name's Tooba. I am the CEO of Amoeba AI. We're building an AI data scientist for go-to market teams.
Fal: Okay. All right. A data scientist for go-to market team, so basically bringing in a data scientist within your team, acting as a person on your team.
So what does that data scientist do? What's their function?
Tooba: So essentially similar to a human data scientist, we sit kind of in a central function, so you can do anything and everything, but we primarily focus on giving you the right things to pay attention to in your data. And they're telling you what you should do with them.
But it's very open-ended. You can ask it any kind of questions like you would with a normal data scientist, or you can ask it to conduct like long-term research, over trends, but the idea there is that you don't have to worry about what you're missing in your data. The data scientist looks after that.
Fal: So, I'm a career coach and so I talk to people at all levels, beginning, middle, and late. As a CEO, you're in a position to hire. What are you looking for in employee now?
What advice would you give to people, especially older workers who are trying to find their way in this industry? Any advice you would give in terms of how they could retool themselves, what they should be focusing on, what the attitude is? What kind advice?
Tooba: I think attitude wise, foundationally you should be your own friend of learning.
I always say that. Because, there's new stuff coming out every day. So if the only skill we teach ourselves is like how to learn to use new tools, how to experiment, how to play with, that's what excites me about now.
Fal: Make friends with the data scientist agent.
Tooba: Yeah, exactly. Yeah. I see. Like I'm not a marketer, but I'm a marketer's friend and I'm designing a marketer's friend, right?
So the more people are in a mindset of I wanna get my hands on it to play with it, that mindset takes you a lot of places in this world. And then the second thing is, humility needs to be present when you do these things because the minute you tell yourself, I know everything is the minute you fall flat.
I think specifically for new folks who are coming in and starting their own companies, create your own blueprint of what you're going to do. I see a lot of advice out there about structuring your day in a certain way or dedicating time to your family, and I don't say don't do that, but I'm saying you have your own blueprint of how you do that.
My blueprint right now is for the next couple of months, I'm obsessively about my company. And I've done the work of setting up my family and my friend groups, et cetera, in a way where they can support me outside of that. But I will never apologize for the fact that I don't get a five 5:00 AM and work out every morning.
Because my flavor is, I wake up obsessively thinking about this problem. I go to sleep obsessively thinking about this problem. So don't be afraid to have your own flavor of what this means for you, because I think that's what people probably appreciate the most.
Fal: Yeah. So self-analysis, understanding who you are
Tooba: Yeah.
Fal: To plot the way of
Tooba: Don't lose yourself in the process. Because there, there are a lot of examples out there of people who have made it in their own ways, right? So it's easy to say there's a blueprint, and I should follow it.
But there's a balance between learning from what people have done versus just like conforming into those people.