Mike

Chartered Accountant now in AI

[Mike @M1keJW]

@kimmonismus I’m finding the same thing, it also grounds its answers in random insignificant previous conversations way too much

[Mike @M1keJW]

@Ameer_Kotecha My suspicion is that he’s intentionally on the charm offensive initially to build some grace & political capital before tackling the scrutiny that will come with policy announcements

[Mike @M1keJW]

@StockSavvyShay I think it’s such a obvious play. Big corporations just won’t sign off agentic ai making up its own workflows and then executing it with no real auditable log of what it has done.

There has to be a control layer that is reviewable by risk management teams and then signed off.

[Mike @M1keJW]

@jzrdan Completely agree! That application layer is crucial for AI agents to be able to operate in a way that is both governable and informed enough by the structured context that the platform provides

[Mike @M1keJW]

@DudeWhoInvests I am really bullish on this stock long term. The integration in large organisations and structuring of data & process is crucial for GenAI agents to operate effectively - not dissimilar to Palantir’s product (albeit less specialist)

[Mike @M1keJW]

@PolitlcsUK I don’t know why this is considered such a bad thing - surely you want the brightest and best running the country?

Bigger issue is that only ~5 of them have ever worked somewhere with client accountability - the rest have spent their whole career in/around politics and advocacy

[Mike @M1keJW]

@michaeljburry Is there any indication as to the timeframe of these commitments?

[Mike @M1keJW]

@CliffordSosin Interesting, feels like robotics could solve the access problem (closing the loop without a human running the experiment), but not the time problem. Then the cap becomes how many experiments you can run in parallel; so, energy, compute, materials etc.

[Mike @M1keJW]

@GaryMarcus Agree - I think there’s value in integrated software that governs workflows and authorisations both in the era of LLMs and AGI.

[Mike @M1keJW]

@HarryStebbings Hugely skeptical of this. The extrapolation of this is that each and every company has a slightly unique salesforce variant that is built by one person. Each then will need updating and support. It might work for some, but largely completely infeasible

[Mike @M1keJW]

@chamath What happens to OpenAI and Anthropic in this situation? Google, xAI and Meta are surely the only survivors of this?

[Mike @M1keJW]

@jawwwn_ It’s a great product but the skeptic in me thinks that scaling it relies on a specific type of smart person who is able to embed reflect the organisation in the ontology. They’re hard to find.

[Mike @M1keJW]

Open-source models keep closing the capability gap and compute is less of a moat than assumed. Once capability is good enough, competition forces the model layer toward commodity pricing. The value that’s left will sit at the context and control layer around the model.

[Alex Finn @AlexFinn]

I was wrong. I said we were a year away from Fable 5 on our desk.

That day is today

An open model BETTER than Fable 5 in some benchmarks just dropped

Better than ChatGPT 5.6 on FrontierSWE. Better than Fable 5 on Automation Bench

This fundamentally changes the AI race forever

If people can start running Fable 5 on their desk, unlimited and for free, they are not going to pay thousands for subscriptions

Now let's be clear: will the hardware you need to run Kimi K3 be obtainable for most? No it won't

It will require multiple very expensive Nvidia chips or a bunch of Mac Studio 512gbs

But look at the slope, not the Y intercept. Over the last year local AI has become SIGNIFICANTLY more efficient and required way less compute

The smartest brains in the world are all attacking the compute problem as hard as they can

This is another step in that direction. And within a year or two, you'll be able to run this on a Mac Mini

Local AI has arrived, and it's not going anywhere

[Mike @M1keJW]

@0xAlaric It’s becoming increasingly obvious that in a business context, the race won’t be won my frontier models, but rather cheap good enough models.

Context provision on the application layer is clearly crucial for this also.

[Mike @M1keJW]

@gsivulka Really interesting read! The problem with $PLTR is that it relies on a certain type of smart person to be able to understand a company's system and encode it for the ontology. This opens the door for companies like $NOW and $SAP which already have a foothold in firms' process.

[Mike @M1keJW]

@uncover_ai Great clip! Would be great to see when the talks are from too

[Mike @M1keJW]

@moving_charlie You see it all over Porto. Owners with no incentive/ability to invest in the property, who let it rot away as their tenants live in it.

[Mike @M1keJW]

@jrouldz I really like the stock as a longer push.

For me, it’s the integration with large corporations workflows and more importantly, the fact that AI agents require good structured data and authorisations which is something that Service Now provides.

[Marc Andreessen 🇺🇸 @pmarca]

Five dimensional chess doesn’t exist. Everyone is furiously improvising all the time. The future is utterly uncertain.