Waymo picks Singapore for its first Southeast Asian robotaxis

Share
Waymo picks Singapore for its first Southeast Asian robotaxis

Waymo's international expansion added a fourth flag today, and this one lands in a country that has already legislated for the drivers it might displace.


Waymo will launch a commercial robotaxi service in Singapore by 2028, its first market in Southeast Asia, working with the Ministry of Transport and the Land Transport Authority. An initial fleet of all-electric Jaguar I-PACE vehicles arrives in the coming months, mapping and validation follow in 2027 — with trained autonomous specialists driving manually at first so the Waymo Driver learns local road geometry and monsoon weather — and the company says it is working toward fully autonomous public ride-hailing in 2028. Rides will be bookable through the Waymo app.

Singapore becomes Waymo's fourth market outside the United States, after London, Tokyo (targeting a 2027 launch with Nihon Kotsu and GO) and Munich. It is also the first where the government has pre-funded the transition: Singapore's parliament passed a manpower transition package in July 2026, and LTA said today it will help drivers retrain and move into roles the AV sector creates, adding that it "will not allow autonomous vehicle deployment to run ahead of our ability to retrain and support affected drivers." Transport minister Jeffrey Siow welcomed Waymo as the country's newest AV operator.

The commercial logic is straightforward. Waymo told Singapore it wants to bridge first-and-last-mile gaps around transit hubs, and cited its own data that more than a third of San Francisco riders are picked up or dropped off near a transit station — the pitch that matters in a city-state where the MRT already works. Waymo has around 4,000 vehicles across roughly 15 US cities, was valued at $126 billion in February when it raised $16 billion, and arrives in a market where Grab and WeRide already run autonomous shuttles, so it is buying position rather than opening a frontier.

Our read: this is the first Waymo expansion where the labour policy exists before the fleet does — and that is the experiment worth watching, not the cars. Every other market has fought the displacement question after deployment, in driver protests and emergency hearings; Singapore is testing whether a retraining package announced in advance actually holds when the vehicles scale. The risk is asymmetric. If the AVs underperform, the transition money was wasted. If they work, Singapore has a template other governments will copy, and Waymo gets a regulator that has already decided the answer to the question that stalls it at home. We covered the Japan half of this build-out earlier this week — Waymo will run Tokyo's first driverless taxis in 2027.


Alibaba added an omni-modal model to its Qwen 3.8 line, aimed at audio and video understanding rather than text. The model, listed in Alibaba Cloud Model Studio as qwen3.8-omni-flash, takes text, image, audio and video input and returns text, and the documentation positions it for audio and video understanding and content analysis — the workload where a fast, cheap multimodal pass beats a frontier model on price. Developers discussing the release on Hacker News noted the framing puts its audio-visual performance near Gemini 3.8 Flash, though the company's own evaluation page for the model is thin on numbers.

The interesting part is where it sits in the family. Alibaba's Qwen 3.8-Flash went after agentic coding and long-horizon office work with an architecture aimed at cost — the company says it trained at less than a ninth of what its previous generation cost — while the omni variant goes after the modality mix that video, meeting and moderation pipelines actually need. Two different races, one price war: Alibaba is now competing on being the cheap default for whatever an application happens to ingest.

What to watch: whether Waymo's 2027 validation phase in Singapore stays on schedule, and whether LTA's driver-support pledge survives the first commercial quarter.

Should governments budget for worker displacement before robotaxis deploy, or does that just buy the operator's political cover? Tell us in the comments.

Read more

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

Nvidia-backed Firmus pulls its $30 billion IPO after demand collapses

The biggest AI-infrastructure listing of the year didn't get delayed — it got pulled entirely, after two price cuts in two days found no takers. Nvidia-backed data-center operator Firmus withdrew its Australian IPO on Friday, abandoning what would have been the country's second-largest listing ever after slashing the offer price twice in 48 hours. The company had been marketing shares at A$11 each to raise about $5 billion at a valuation near $30 billion — roughly three times the $10.5 billion

AI inference is redrawing the storage hierarchy

AI inference is redrawing the storage hierarchy

The money has followed GPUs for years, but at GMIF 2026 in Shenzhen the argument was about everything behind them — and the numbers backing it are getting hard to ignore. China's 140 trillion daily token calls are forcing a redesign of the storage stack. Now in its fifth year, the Global Memory Innovation Forum gathered storage makers, analysts and chip designers around one theme: inference, not training, is now the workload that dictates hardware. China's daily token calls had already passed

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

Cloudflare launches Clef-omni, cuts Clef-flash price by 58%

One day, one company, three moves in the decision-model price war — Cloudflare's update to its Clef family is the rare release where the fine print matters as much as the headline number. Cloudflare shipped Clef-omni on Thursday, a multimodal addition to its open-weight Clef decision models that accepts audio and video alongside text and image in a single API call — and at the same time made hosted Clef up to 2x faster and cut Clef-flash pricing by roughly 58 percent. Clef-omni takes WAV or MP

Meta turned down Amodei's personal plea for compute

Meta turned down Amodei's personal plea for compute

The chip hunt is back in the headlines — and today's edition runs from a boardroom ask at the very top of the AI industry to the economics of putting robot drivers in truck cabs. Dario Amodei personally approached Meta earlier this year to source more compute for Anthropic, and Meta declined — according to the Wall Street Journal. The report lands inside the Journal's larger piece on how desperately the industry is hunting for computing power, and it is the kind of detail that only surfaces wh