Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more

Artificial intelligence did not begin with ChatGPT. In a December 2024 commentary, Marc Solomon traces a long path from early AI research to today’s generative tools, then argues that organizations should focus less on producing more content and more on finding useful information that helps people make decisions. His account is an enterprise-focused perspective, not a full history of AI or proof of current business returns.

AI’s past: a long road to the generative AI boom

Solomon’s compressed history begins with the 1956 Dartmouth workshop on “thinking machines,” organized by John McCarthy, then a Dartmouth mathematics professor. He sketches later research and investment, IBM Deep Blue’s 1997 chess match against Garry Kasparov, the arrival of consumer voice assistants, and ChatGPT’s launch in 2022. The point is that the technology has a much longer history than its recent surge in public attention.

That outline is a high-level account in Solomon’s commentary, rather than a detailed chronology with evidence for every milestone. It is most useful as context for his argument: generative AI is a conspicuous new phase in a field that predates it by decades.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

AI at work today: promise and pressure

Solomon describes organizations’ interest in generative AI in terms of productivity, streamlined operations, and cost efficiencies. In cybersecurity, he notes that AI-related tools have been used for decades to support threat detection, response, and system security. These are broad descriptions of intended applications—not quantified evidence that a particular deployment produces net benefits.

The challenge, in his view, is that impressive outputs and intense attention can outpace the work of identifying a valuable business problem. An organization can add complexity or choose a poor-fit project if it adopts AI mainly to keep up with hype or demonstrate activity.

Is generative AI delivering business value?

The figures Solomon cites suggest that many U.S. generative AI decisionmakers expected returns to take time. He reports that, in Forrester’s Q2 2024 AI Pulse Survey, 49% expected their organization to realize ROI within one to three years, while 44% expected it within three to five years. These are expectations reported through Solomon’s article, not measured outcomes or a current survey result.

The practical distinction is between a compelling demonstration and durable organizational value. A pilot can help test whether a use case fits a business need; it does not, by itself, establish that a wider rollout will pay off. Solomon’s advice is to link adoption to business goals and assess returns over a multi-year horizon instead of demanding immediate proof or scaling simply because the technology is prominent.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

What might come next: Solomon’s “SynthAI” idea

Solomon proposes “SynthAI” as a possible next direction: systems that sift large collections of information, surface relevant themes, and help people make decisions. This is his suggested term and vision, not an established industry category or a confirmed forecast. The opportunity he describes is greatest where information is abundant, difficult to review manually, and noisy enough that important signals are hard to find.

Approach Primary emphasis Potential role in a decision
Generative AI Creating content or responses Produce material that people can review and use
Synthesis-oriented AI (“SynthAI” in Solomon’s proposal) Filtering and synthesizing existing information Bring relevant material and themes forward for human judgment

This is a conceptual contrast drawn from Solomon’s argument, not a measured comparison of system performance. Synthesis still requires people to judge whether surfaced information is relevant and reliable; reducing a large collection is useful only when it improves the decision that follows.

Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Support on Ko-Fi

How companies can adopt AI without chasing hype

  1. Start with a business problem. Define the operational or security need before choosing a tool, and specify what a useful result would look like.
  2. Test a bounded use case. Use a pilot to learn whether the system fits the task and where human review or other guardrails are needed.
  3. Evaluate the result against the goal. Look beyond a convincing output: determine whether the use case meaningfully supports the intended work and merits further investment.
  4. Plan for a realistic return horizon. Treat ROI as a multi-year question where appropriate, rather than assuming adoption should produce immediate financial gains.
  5. Scale deliberately. Expand only when the business fit and safeguards justify it, not because visibility or competitive pressure makes deployment feel urgent.

Solomon’s closing recommendation was written with 2025 and beyond in view: “But I would advise on AI with caution or AI with guardrails and a clear focus on how to return a multi-year ROI in 2025 and beyond.” The year belongs to his original 2024 framing; it should not be read as a new forecast for 2026.

Read Marc Solomon’s full SecurityWeek commentary (December 12, 2024).

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.