After a great deal of anticipation, Sam Altman and the OpenAI team finally released ChatGPT-5 at the OpenAI Summer Update yesterday. While there’s been a fair bit of subtle hinting that the launch was imminent, the last indication was a little more on the nose, shall we say – OpenAI published a Tweet stating simply, “LIVE5TREAM THURSDAY 10AM PT”. Sure, they didn’t explicitly state that the live stream was all about GPT-5, but I’d like to think that, while some people may have assumed some silly intern made a big old career-ending typo, the majority of the tech community took careful note of that curious “5”.
Now, there’s a lot to be said about the Summer Update – from all of the exciting new features and upgrades of GPT-5 to the fact that I am now about 85% certain that Sam Altman is an actual robot. They announced big changes in coding capabilities, impressive new applications in health and medtech as well as the promise of far fewer instances of hallucinations – all in all, some pretty monumental steps forward.
So, for those who had better things to do yesterday than be simultaneously impressed and perplexed – and also kind of lulled into a bizarre trance by Altman’s carefully scripted monologues and the uncomfortable pseudo-off-the-cuff dialogue between OpenAI engineers and execs – we’ve put together a little explanation of what you missed. Because, as much as we may make jokes about the employees’ cringeworthy presentation skills – no shade whatsoever, they’re clearly insanely skilled and experienced technical experts – the work they’ve done on GPT-5 is pretty impressive, and it seems like a pretty darn big move towards industry-wide ambitions to achieve GenAI.
GPT-5: The Highlights
For those who want a quick snapshot of what’s new and what was announced yesterday, here’s a quick summary fo the main updates, changes and new features of OpenAI’s brand new ChatGPT-5.
Core Upgrades to the Model
These are the primary upgrades you can expect shifting from GPT-4 to GPT-5:
- Unified AI System: GPT-5 can easily, quickly and efficiently switch between “fast/basic,” “deep reasoning” and “mini/nano” modes depending on the complexity of your request.
- Sharper reasoning: Significant improvement in multi-step logic, problem solving and context analysis.
- Far Fewer Hallucinations: GPT-5 will hallucinate less than ever before, and it’s also less “deceptive”.
- State-of-the-Art Performance Across a Variety of Domains: It’s especially strong in coding, maths, writing, health advice and visual understanding.
- Improved Accuracy: More reliable, factual responses with better source grounding.
These are the promises, and they’re pretty monumental.
New UX Features
User experience is a significant component of how much people like using AI models, so these improvements and new UX features could go a long way in making GPT-5 a new favourite:
-
Personalities: Choose from different conversational “styles” for the AI (e.g., friendly, concise, professional).
-
Customisable interface: New themes and layout options for personalisation.
-
Voice Upgrades: Far more natural-sounding voices, faster response in voice mode and better back-and-forth flow.
-
Tool Integrations: Direct connection to Gmail, Google Calendar and other services for scheduling, email drafting and workflow help.
These new UX features sound cool, and to be honest, judging from the demos, they look pretty cool too.
Specialised Capabilities
This is undoubtedly where most people were focused, and for good reason. These changes are pretty big, and OpenAI’s focus – especially their kind of surprising emphasis on health-related issues – was fascinating.
- Enhanced Coding: It’s cleaner, there’s more modular code generation, better debugging, improved adaptation between languages and stronger context memory for multi-file projects. It’s impressive, there’s no way around it.
- Advanced Healthcare Guidance: Smarter, more context-aware health information but it came with very, very clear disclaimers that this is not a replacement for medical professionals.
- Multimodal Improvements: Better image understanding and integration with text-based reasoning.
These three points are pretty important, and the focus is intriguing – coding is no surprise, but health (especially their interview with users who personal use of GPT-5 during a particularly difficult personal medical situation) was not on my bingo card for yesterday’s presentation.
Access and Availability
So, the one thing everyone wants to know – what about access? How good is the free version going to be and what are the options for upgrades? Well, the unpaid versions of GPT-5 are actually pretty generous, in my opinion, which I’m sure will be good news to plenty of users.
- Free Access for All: GPT-5 available to all ChatGPT users, with usage limits for free accounts, of course.
- Pro Tier Advantages: Pro subscribers get extended usage, priority speed and access to the most advanced “pro reasoning” mode – that’s the big, exciting part!
- Mini/Nano Versions: Lightweight variants for mobile and low-resource situations, with faster responses.
All in all, OpenAI has made access to GPT-5 pretty inclusive – I mean, anyone can use GPT-5 for free, and even though their are limits on how much you can use it, that’s pretty useful.
AI Safety
Unsurprisingly, given the fact that safety and security have always been a massive topic in AI and only seem to be becoming more prominent, the Summer Update did focus on the issue with a comprehensive explanation and a decent demo. Here are the highlights:
- Inputs Are Safer and More Aligned: Reduced bias, fewer misleading answers and less “sycophancy” (randomly agreeing without thinking).
- GPT-5 Framed As a “Proactive Thought Partner”: Clear nods towards moving closer toward artificial general intelligence (AGI) but not AGI yet.
- Refined Memory: GPT-5 keeps relevant context without bloating responses or repeating information.
All important, with the obvious emphasis on the specific reference to a “proactive thought partner” and not making any wild claims about AGI.
Main Takeaways from OpenAI’s Summer Update
So, this all sounds cool, and futuristic and oh-so-impressive, but what do the experts think about GPT-5? Has it lived up to the hype? I spoke to Paul Armstrong from the TBD Group, Founder and creator of ‘What Did OpenAI Do This Week?’, to find out what he has to say about yesterday’s updates, and all in all, his response was equally optimistic and measured. In summary, Armstrong definitely thinks that GPT-5 is a step up, but, as always, he’s cautious about some potentially overreaching statements, and I have to agree.
According to Armstrong, “GPT‑5 is a systems-level upgrade, not just another model release. The real advance isn’t in language but in orchestration. The router architecture allows it to manage its own workload, deciding when to respond quickly or reason more deeply. That shifts user expectations. You’re not picking the model anymore. The model is picking its own brain.
Still, claims of “PhD-level intelligence” are misleading. GPT‑5 is better tuned, more honest when it fails, and slightly more cautious. But it is still a static, probabilistic output machine. It doesn’t think, it predicts. The main difference is that now it knows when to pull back instead of bluffing.”
One of the major questions most people (myself included) have about GPT-5 is simple – how does it compare to GPT-4? While GPT-4 may have been the shiny new thing not too long ago, there’s no doubt about the fact that GPT-5 outdoes it on loads of different fronts.
“GPT‑4 was erratic and brittle under pressure. GPT‑5 is more stable, less prone to collapse when prompted aggressively or asked to perform multi-step logic. It avoids the embarrassing failures that defined GPT‑4’s limitations in professional use. The leap isn’t in cognitive architecture, it’s in constraint management. GPT‑5 reduces hallucination, trims sycophancy, and defaults to silence when uncertain. That makes it more useful in practice. The underlying model is still static and narrow in scope, but the packaging now suits operational environments where reliability matters more than novelty,” says Paul.
All in all, he thinks, “users are going to love it,” and it’s hard to contest this one.
Now, we’ve got the gist of general reactions to GPT-5, I’d like to think. But, what about two of the main bits – in my eyes, two standout features and points of emphasis from the live stream – that is, applications in health and coding? I asked Armstrong about both, and here’s what he had to say.
What Does GPT‑5’s Increased Coding Capability Mean for AI?
“GPT‑5 shifts coding from execution to intent. You’re no longer writing code, users are simple issuing design directives. The demos were slick and aimed at lots of different levels which was smart. OpenAI showed that it handles more than just syntax and logic, but also stylistic and architectural elements. Now that’s an inflection point; prompting becomes programming will open doors to a lot more people and side hustles whilst simultaneously changing the role of the developer.
Coding is not unvaluable by any means, but what’s clear is the direction of travel. These models will just get better and better, and developers will likely learn how to supervise systems, not script them.”
What Do You Think of GPT‑5 in the Medical and Health Context, Especially Given the Privacy and Confidentiality Concerns?
“Health was the interesting push in the presentation. Mentioned early and all the way through, with an emotional element you don’t often see in these events. All that said, deploying GPT‑5 in health contexts is premature. The model performs well on benchmarks, but benchmarks aren’t medicine. The model doesn’t reason diagnostically, and it is imitating authority. That’s not competence, that’s pattern-matching with confidence and will be problematic.
This said, there’s a lot to like with how it can empower people to have better exchanges with doctors before, during and after appointments. Still, the dangers are clear: until GPT‑5 can be independently audited, verified, and governed in clinical contexts, the use cases should be considered helpful at best. However, again, you can see the direction of travel, the model will only get smarter and more attuned to answer questions without adding ambiguity or changing the doctor/patient trust relationship. ”
That’s a Wrap on GPT-5 (for Now)
Well, there you have it – OpenAI’s highly-anticipated ChatGPT-5 model is now out in the world (well, it’ll be totally out there by the end of next week), ready to be tested and enjoyed in a plethora of new and exciting applications.
As always, there are a few things that may be a little overhyped and a handful of statements that give rise to a “well, maybe not quite” kind of response, but all in all, I’d say GPT-5 looks pretty darn good and it’s going to interesting to see not only what people do with it, but where OpenAI plans to go after this.
So, get testing (it’s free!) and let us know what you think about it.
