Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now
Here’s an analogy: Freeways didn’t exist in the U.S. until after 1956, when envisioned by President Dwight D. Eisenhower’s administration — yet super fast, powerful cars like Porsche, BMW, Jaguars, Ferrari and others had been around for decades.
You could say AI is at that same pivot point: While models are becoming increasingly more capable, performant and sophisticated, the critical infrastructure they need to bring about true, real-world innovation has yet to be fully built out.
“All we have done is create some very good engines for a car, and we are getting super excited, as if we have this fully functional highway system in place,” Arun Chandrasekaran, Gartner distinguished VP analyst, told VentureBeat.
This is leading to a plateauing, of sorts, in model capabilities such as OpenAI’s GPT-5: While an important step forward, it only features faint glimmers of truly agentic AI.
AI Scaling Hits Its Limits
Power caps, rising token costs, and inference delays are reshaping enterprise AI. Join our exclusive salon to discover how top teams are:
- Turning energy into a strategic advantage
- Architecting efficient inference for real throughput gains
- Unlocking competitive ROI with sustainable AI systems
Secure your spot to stay ahead: https://bit.ly/4mwGngO
“It is a very capable model, it is a very versatile model, it has made some very good progress in specific domains,” said Chandrasekaran. “But my view is it’s more of an incremental progress, rather than a radical progress or a radical improvement, given all of the high expectations OpenAI has set in the past.”
GPT-5 improves in three key areas
To be clear, OpenAI has made strides with GPT-5, according to Gartner, including in coding tasks and multi-modal capabilities.
Chandrasekaran pointed out that OpenAI has pivoted to make GPT-5 “very good” at coding, clearly sensing gen AI’s enormous opportunity in enterprise software engineering and taking aim at competitor Anthropic’s leadership in that area.
Meanwhile, GPT-5’s progress in modalities beyond text, particularly in speech and images, provides new integration opportunities for enterprises, Chandrasekaran noted.
GPT-5 also does, if subtly, advance AI agent and orchestration design, thanks to improved tool use; the model can call third-party APIs and tools and perform parallel tool calling (handle multiple tasks simultaneously). However, this means enterprise systems must have the capacity to handle concurrent API requests in a single session, Chandrasekaran points out.
Multistep planning in GPT-5 allows more business logic to reside within the model itself, reducing the need for external workflow engines, and its larger context windows (8K for free users, 32K for Plus at $20 per month and 128K for Pro at $200 per month) can “reshape enterprise AI architecture patterns,” he said.
This means that applications that previously relied on complex retrieval-augmented generation (RAG) pipelines to work around context limits can now pass much larger datasets directly to the models and simplify some workflows. But this doesn’t mean RAG is irrelevant; “retrieving only the most relevant data is still faster and more cost-effective than always sending massive inputs,” Chandrasekaran pointed out.
Gartner sees a shift to a hybrid approach with less stringent retrieval, with devs using GPT-5 to handle “larger, messier contexts” while improving efficiency.
On the cost front, GPT-5 “significantly” reduces API usage fees; top-level costs are $1.25 per 1 million input tokens and $10 per 1 million output tokens, making it comparable to models like Gemini 2.5, but seriously undercutting Claude Opus. However, GTP-5’s input/output price ratio is higher than earlier models, which AI leaders should take into account when considering GTP-5 for high-token-usage scenarios, Chandrasekaran advised.
Bye-bye previous GPT versions (sorta)
Ultimately, GPT-5 is designed to eventually replace GPT-4o and the o-series (they were initially sunset, then some reintroduced by OpenAI due to user dissent). Three model sizes (pro, mini, nano) will allow architects to tier services based on cost and latency needs; simple queries can be handled by smaller models and complex tasks by the full model, Gartner notes.
However, differences in output formats, memory and function-calling behaviors may require code review and adjustment, and because GPT-5 may render some previous workarounds obsolete, devs should audit their prompt templates and system instructions.
By eventually sunsetting previous versions, “I think what OpenAI is trying to do is abstract that level of complexity away from the user,” said Chandrasekaran. “Often we’re not the best people to make those decisions, and sometimes we may even make erroneous decisions, I would argue.”
Another fact behind the phase-outs: “We all know that OpenAI has a capacity problem,” he said, and thus has forged partnerships with Microsoft, Oracle (Project Stargate), Google and others to provision compute capacity. Running multiple generations of models would require multiple generations of infrastructure, creating new cost implications and physical constraints.
New risks, advice for adopting GPT-5
OpenAI claims it reduced hallucination rates by up to 65% in GPT-5 compared to previous models; this can help reduce compliance risks and make the model more suitable for enterprise use cases, and its chain-of-thought (CoT) explanations support auditability and regulatory alignment, Gartner notes.
At the same time, these lower hallucination rates as well as GPT-5’s advanced reasoning and multimodal processing could amplify misuse such as advanced scam and phishing generation. Analysts advise that critical workflows remain under human review, even if with less sampling.
The firm also advises that enterprise leaders:
- Pilot and benchmark GPT-5 in mission-critical use cases, running side-by-side evaluations against other models to determine differences in accuracy, speed and user experience.
- Monitor practices like vibe coding that risk data exposure (but without being offensive about it or risking defects or guardrail failures).
- Revise governance policies and guidelines to address new model behaviors, expanded context windows and safe completions, and calibrate oversight mechanisms.
- Experiment with tool integrations, reasoning parameters, caching and model sizing to optimize performance, and use inbuilt dynamic routing to determine the right model for the right task.
- Audit and upgrade plans for GPT-5’s expanded capabilities. This includes validating API quotas, audit trails and multimodal data pipelines to support new features and increased throughput. Rigorous integration testing is also important.
Agents don’t just need more compute; they need infrastructure
No doubt, agentic AI is a “super hot topic today,” Chandrasekaran noted, and is one of the top areas for investment in Gartner’s 2025 Hype Cycle for Gen AI. At the same time, the technology has hit Gartner’s “Peak of Inflated Expectations,” meaning it has experienced widespread publicity due to early success stories, in turn building unrealistic expectations.
This trend is typically followed by what Gartner calls the “Trough of Disillusionment,” when interest, excitement and investment cool off as experiments and implementations fail to deliver (remember: There have been two notable AI winters since the 1980s).
“A lot of vendors are hyping products beyond what products are capable of,” said Chandrasekaran. “It’s almost like they’re positioning them as being production-ready, enterprise-ready and are going to deliver business value in a really short span of time.”
However, in reality, the chasm between product quality relative to expectation is wide, he noted. Gartner isn’t seeing enterprise-wide agentic deployments; those they are seeing are in “small, narrow pockets” and specific domains like software engineering or procurement.
“But even those workflows are not fully autonomous; they are often either human-driven or semi-autonomous in nature,” Chandrasekaran explained.
One of the key culprits is the lack of infrastructure; agents require access to a wide set of enterprise tools and must have the capability to communicate with data stores and SaaS apps. At the same time, there must be adequate identity and access management systems in place to control agent behavior and access, as well as oversight of the types of data they can access (not personally identifiable or sensitive), he noted.
Lastly, enterprises must be confident that the information the agents are producing is trustworthy, meaning it’s free of bias and doesn’t contain hallucinations or false information.
To get there, vendors must collaborate and adopt more open standards for agent-to-enterprise and agent-to-agent tool communication, he advised.
“While agents or the underlying technologies may be making progress, this orchestration, governance and data layer is still waiting to be built out for agents to thrive,” said Chandrasekaran. “That’s where we see a lot of friction today.”
Yes, the industry is making progress with AI reasoning, but still struggles to get AI to understand how the physical world works. AI mostly operates in a digital world; it doesn’t have strong interfaces to the physical world, although improvements are being made in spatial robotics.
But, “we are very, very, very, very early stage for those kinds of environments,” said Chandrasekaran.
To truly make significant strides requires a “revolution” in model architecture or reasoning. “You cannot be on the current curve and just expect more data, more compute, and hope to get to AGI,” she said.
That’s evident in the much-anticipated GPT-5 rollout: The ultimate goal that OpenAI defined for itself was AGI, but “it’s really apparent that we are nowhere close to that,” said Chandrasekaran. Ultimately, “we’re still very, very far away from AGI.”
[ad_2]
Source link
