AI for individuals, businesses and enterprise
Key features of prominent artificial intelligence services
During my time at the National Artificial Intelligence Centre working to grow Australia’s AI ecosystem I was immersed in the landscape of AI product and services for the adoption of AI capabilities in Australian businesses. Similar to many advanced economies, 99% of businesses in Australia are small to medium sized companies, employing approximately 65% of the workforce.
Although generative AI has entered the mainstream in recent years and it’s likely being used by employees in small to medium sized businesses, it’s unlikely these organisations are large enough to employ AI experts to advise on the appropriateness of tools currently in use, or a more ideal AI strategy. As such, this week’s Brief provides a strategic overview of prominent categories of AI services to help inform decisions on how to best utilise these technologies.
Tina
★
AI for individuals, businesses and enterprise
A short guide on the current mainstream generative AI and AI service categories.
To assist individuals as well as employees and decision makers at small to medium sized businesses considering AI in the workplace, the table below identifies key features and differences between consumer generative AI, generative AI offers for business and enterprise AI solutions.
Data protection
If using any of the generative AI platforms without a team or business plan, make sure to opt out of the use of your content for training data so that it isn’t used in future models. You’ll find this option in the settings under privacy.
To reduce the risk of breaching data protection and privacy policy and legislation (e.g. GDPR in Europe, the Privacy Act in Australia), unless you’re working with an AI service provider that guarantees data protection, it’s safest to avoid ever placing any personal identifiable information into prompts (names, addresses). Even if subscriptions specify that data is not used for training or other purposes or you specifically opted out, the Terms of Service of AI providers works in their favour. If there is a bug or glitch and data input is accidentally fed into their model, they are not liable.
Looking ahead
Large language model + industry specific aggregated data >>> specialist capabilities for small and medium businesses
There is potential for significant growth in the business or teams middle ground options between enterprise AI and consumer level generative AI. For example, startups that offer small to medium sized businesses generative AI capabilities with enterprise AI features focused on specific industries or professional roles and provided as a cloud based service. This could assist people working in consultancies or small businesses that undertake repeatable processes or report writing based on their industry expertise, but do not have an organisation big enough to fine tune a large language model with their knowledge or company data (e.g. past work, reports).
With team subscriptions or third party applications it’s already possible to have a large language model, such as ChatGPT, draw from information uploaded and follow style guides. However, better quality outputs would be possible with aggregated data. That said, the problems of large language model hallucination and black box decision making remain, so assisting people who have the expertise to scrutinise outputs is still key.